As is well known, hydraulics is used to transfer energy by means of a pressurized medium, wherein a pressurized volume flow produced by pumping means is controlled by valve means and utilized in actuator means, such as cylinders and engines, to produce a linear motion, a force, a moment, or a rotational motion. Known pressurized media include hydraulic oil, pressurized air and water or water-based hydrau- lic fluids (HFA, HFB, HFC, HFD).

For the control of cylinders and engines, proportional control valves are used in a known manner, one example being presented in US patent 5,785,087. The disclosed valve is a slide-type valve, in which the posi- tion of the slide is controlled by means of a proportional magnet to choke the pressurized medium and simultaneously to control the quan- tity and direction of the volume flow. The valve is controlled by an external electric set signal which is proportional to the desired volume flow. The valve is used to provide stepless control of the volume flow.

It is an aim of the present invention to present an alternative method to control the volume flow particularly in connection with actuators, which method also provides energy savings. The invention provides consid- erable advantages to conventional proportional control and eliminates problems therein. The control system according to the invention is pre- sented in claim 1. The method according to the invention is presented in claim 12.

In particular, the control system is suitable for controlling the rate of an actuator, but it can also be applied in position control when continu-

ously monitoring a desired or set position dependent on the moment of time.

Systems are known for the stepless control of a volume flow imple- mented by means of separate simple valves. In the system, the control to different volume flow levels is implemented with valves which only have an open position and a closed position (so-called on/off valve, 2/2valve). A 4/3 valve (4 ports and 3 positions) implemented with separate valves is, in view of some of its features, even more versatile than a proportional 4/3 valve.

As on/off components, it is possible to use very inexpensive, easily available valves where it is possible to use simple structures with no problems related to wearing in connection with the sealing or the throt- tling of the volume flow. On the other hand, even short lifetimes are allowable for such components, because of their inexpensiveness and simple installation. Another advantage is their resistance to impurities.

One example of an on/off system that can be coupled to the pressure or return side of a cylinder is presented in US patent 2,999,482. US patent 5,313,871, in turn, discloses the control of an on/off system on the pressure side, to transfer the cylinder to a desired extent. US pat- ent 4,590,966, in turn, discloses one version of the on/off system, i. e. a kind of a digital valve block which is coupled to the pressure and return sides of a valve. A corresponding block type is also presented in US patent 4,518,011.

The number of control steps can be decreased and increased accord- ing to the need and the use, wherein the same valve system can be applied in a more efficient way in different situations. The need of few control steps will also reduce the number of on/off valves, which will cut down the costs. Moreover, the operation of valves, particularly the change in the control, always involves a response time which, in the case of a slide valve, depends on the length of the distance that the slide must move. If necessary, the on/off valves will operate simultane- ously, wherein the delay of operation is substantially constant with all

the changes in the control of the volume flow, because the valves only move between the open and closed positions.

Compared to a slide-controlled valve, another advantage is that flows of two simultaneous connections in the valve system can be controlled separately. In the above-mentioned slide valve, different guide edges of the same slide are used to simultaneously control these flows of the pressurized medium in different directions. The independence of the control in the on/off system makes it possible to reduce unnecessary pressure losses in the flow and thereby also to save energy to a con- siderable extent without affecting the controllability. On the other hand, the functional independence will also increase the ways of using and controlling the 4/3 valve.

In the following description, the invention will be illustrated with refer- ence to the appended drawings, in which Fig. 1 is a schematic diagram showing the characteristic curve of a proportional valve of prior art, Fig. 2 is a schematic diagram showing the characteristic curve of a valve system according to the invention, with respect to one set of valves, Fig. 3 is a block chart showing the structure of a valve system according to a preferred embodiment of the invention, Figs. 4a to 4d show different possibilities to couple the valve system of Fig. 3, Fig. 5 is a schematic view showing a circuit applying an on/off valve system in the control of a cylinder, Fig. 6 shows the rates of the cylinder in the system of Fig. 5 with different combinations of valve openings, Fig. 7 shows the rates of Fig. 6 in an order of magnitude, and

Fig. 8 shows pressure variations in the system of Fig. 5 with different combinations of valve openings.

Figure 1 shows the characteristic curve of a proportional directional valve according to prior art, wherein the transmission capacity of the volume flow Q of the valve is linearly dependent on a control signal V.

The rate of the characteristic curve is also dependent on the pressure difference effective over the valve, that is, the pressure loss in the directional valve. In the presented example, the control signal used is a voltage signal V, but it is also possible to use a current signal/. The level selected for the control signal I is used as a set value VSET, wherein maximum control VMAX corresponds to the maximum of the volume flow QMAx, and, respectively, the other levels of the curve cor- respond to the volume flow QSET, wherein no volume flow is obtained when the control is zero. Normally, the characteristic curve continues as a mirror image on the negative side, wherein VSET may also be negative, wherein the direction of the flow and the position of the valve are changed as well.

In a directional proportional valve, a proportional magnet has an effect on the control slide which is normally centered by means of springs.

The proportional magnet, in turn, is normally a direct current (DC) magnet, whose output is an armature force or a range of motion which is proportional to the level of the input electric control signal. The valve control means are provided on a circuit board which is electrically cou- pled to the valve. To the circuit board, in turn, it is possible to couple a control system for controlling the valve and the larger apparatus. Thus, the control signal with a certain level and obtained from the control system can thus be used to control the volume flow through the valve and further the actuator coupled to the valve. The pressurized volume flow can further be used to control the speed and direction of motion of the cylinder actuator. For example, with a set signal ranging from-10 V to +10 V, the direction of motion of the cylinder (i. e., the position of the valve slide) is determined by the sign, wherein at the limit value of 0 V, the valve is in its closed position and the cylinder is stopped. The set signal can also be, for example, a signal from 0 V to 20 V, wherein the

limit value is 10 V. Figure 1 also shows the graphic symbol of the directional proportional valve and the ports therein.

Figure 2 shows the stepped characteristic curve of one set of valve in an on/off valve system, wherein the volume flow Q is only changed stepwise when the control signal V is changed. The volume flow QSET of the valve system is changed stepwise when the set signal VSET exceeds or goes below the limits set for the steps. Naturally, the con- trollability of the volume flow of the pressurized medium is the better, the smaller the size dQ of the step is in relation to the maximum vol- ume flow QMA) (, that is, the greater the number of valves.

With reference to Fig. 3, an advantageous valve system 100 according to the invention, intended for the control of an actuator, comprises at least ports P, T, A, and B. The inlet port P is arranged for coupling a pressure line in the system to receive a volume flow of a pressurized medium normally from a pump. The outlet port T is arranged to couple a return line in the system, the return line being a tank line or a line with a lower pressure level than the pressure line. The first working port A is arranged to couple an actuator, such as the piston side of a cylinder, in the system. The second working port B is arranged to couple an actuator, such as the piston rod side of a cylinder, in the system. The volume flow is input in the actuator through one working port, and simultaneously the volume flow is received from the actuator through the other working port.

For different directions of motion of the actuator, the system 100 com- prises electrically controllable coupling means 101 which are arranged to form at least a first connection 1 (between the inlet port P and the first working port A) and simultaneously a second connection 2 (between the outlet port T and the second working port B). The cou- pling means 101 are also arranged to form at least a third connection 3 (between the inlet port P and the second working port B) and to form simultaneously at least a fourth connection 4 (between the outlet port T and the first working port A). The different coupling alternatives are also illustrated by means of graphic symbols in Fig. 4c.

The desired coupling alternatives can be implemented, for example, by means of an electrically controlled 4/2 valve, which is an on/off valve V of the slide type as shown in Fig. 3. Alternatively, the connections can be implemented by means of two parallel electrically controlled 3/2 valves. By increasing the position alternatives of the slide valve, it is possible to increase the coupling alternatives of the whole system. The coupling means 101 may thus vary to a great extent, and they can be connected to the system 100 also separately. Thus, in its basic form, the system 100 comprises only the first connection 1 between the inlet port P and the first working port A, as well as the separate second con- nection 2 between the outlet port T and the second working port B. The coupling means 101 are thus coupled to the working ports A and B, when necessary. For example, in connection with the control of an engine, the means 101 will not be needed, if the engine only rotates in one direction.

Further, with reference to Fig. 3, the pressurized medium flowing through the connections 1 to 4 is guided further by valve means 102.

The valve means 102 comprise two sets of valves 103 and 104. Each set comprises several electrically controllable on/off valves which are coupled in parallel and whose mutual pressurized medium throughput capacity is arranged as a stepped series. The minimum capacity of the volume flow of the pressurized medium in the whole valve system 100 is determined by the capacity of the smallest valve in the set, and the maximum capacity by the sum capacity of all the valves. A set 103 (in the port P) is used to control the volume flow in the connections 1 and 3, and a set 104 (in the port T) is used to control the volume flow in the connections 2 and 4. However, either the connections 1 and 2 or the connections 3 and 4 can be controlled simultaneously. The valve sets 103 and 104 can also be assembled in another way to form a desired stepped characteristic curve, wherein the set may comprise several valves of the same size. The valve sets may also be different from each other. The valve set in the example system, in turn, can be easily represented mathematically, and its number of valves is as small as possible in relation to the number of the steps. Furthermore, the control curve achieved with the example system, as shown in Fig. 7, makes a precise, almost linear control possible, particularly at high speeds.

In the presented embodiment of the invention, the sets 103 and 104 are implemented with four electrically controlled 2/2 valves (solenoid controlled on/off valves), as shown in Fig. 3. The mutual pressurized medium volume flow throughput capacities of the valves V1 to V4 are arranged according to table 1. For the control, at least two valves are needed.

In an advantageous embodiment of the invention, as shown in Fig. 3, each set comprises four (number indicated as N) valves 15 to achieve a control step. Thus, the throughput capacity of the first valve (V1) is 0.5 I/min (indicated as Q), and the throughput capacities of the follow- ing valves (V2, V3, V4) are 1.0 I/min, 2.0 I/min and 4 I/min, respectively, that is, according to the calculatory rule 2 (''Q, in which the variable n is the serial number of the valve. Thus, the size dQ obtained for a step is Q, and the maximum number of steps will be 15 (that is, according to the calculatory rule 2N-1), excluding step 0. As a result, the minimum capacity of the system is Q and the maximum capacity is 7.5 I/min (that is, according to the calculatory rule Q (2N-1)). By means of the calcu- latory rules, the required number N of valves can be calculated on the basis of the desired maximum capacity and dQ desired. In a corre- sponding manner, three valves will be needed to provide seven steps, and six valves to provide 63 steps. Typically, the number of valves ranges from 3 to 7. QTOT Valve V1 Valve V2 Valve V3 Valve V4 [I/min] QV1=0,5I/min $QV2=1,0I/min QV3=2,0I/min QV4=4,0I/min 0 0 0 0 0 0,5 1 0 0 0 1,0 0 1 0 0 1,5 1 1 0 0 2,0 0 0 1 0 2, 5 1 0 1 0 3, 0 0 1 1 0 3, 5 1 1 1 0 4,0 (4, 5) 0 (1) o 0 1 4,5 (4, 5) 1 1 0 0 1 5,0 (5, 5) 0 (1) 1 0 1 5,5 5, 5 1 1 1 0 1 6,0 (6,5) 0 (1) 0 1 1 6,5 (6, 5) 1 1 0 1 1 7,0 (7, 5) 0 (1) 1 1 1 1 1 Table 1

To control the throughput capacity of the valve system to a desired level QSET, those valves are opened, whose total capacity QTOT corre- sponds to the desired level and the set signal VSET In table 1, the lev- els thus occur at intervals of 0.5 I/m, and to achieve each level, it is marked on the same line if the valve in question is thus controlled to be open (1) or closed (0). To close the whole connection, all the valves are closed (0). In some situations, the operation of the valve system can be improved, if the smallest valve (V1) is kept open all the time at high volume flows (QSET > QmAx/2). Thus, the number of steps is reduced but may still be sufficient for the control. This situation is indi- cated in parenthesis in table 1.

In Figs. 4a and 4b, graphic symbols are used to illustrate the coupling possibilities of the sets 103 and 104, respectively. The coupling to the port A or B will depend on the position of the coupling means 101. In the presented embodiment, the presented sets are identical, but they may differ from each other as desired. When all the valves V1 to V4 in each set are closed, the connection 1 to 4 can be closed, and if only one valve V1 to V4 in the set is opened, a connection is made, wherein the volume flow passed through the valve system will depend on the valves being open at the time. The controllability of the volume flow Q in the coupling is indicated with a lineation. Considering the overall coupling possibilities of the valve system 100, shown in Fig. 4d, the valve system is found to be capable of providing, in addition to the couplings of a normal 4/3 valve, also couplings in which one connec- tion is closed at a time. This is possible, because the sets 103 and 104 can be controlled separately, if necessary. Furthermore, the separation

also means that pressure losses can also be controlled separately in simultaneous connections.

Further, with reference to Fig. 3, the valve system 100 also comprises electrical control means 105 for controlling the valve means 103,104 by control signals 106. According to an advantageous embodiment of the invention, the control means 105 consist of a programmable micro- controller which is coupled by control conductors to the valves of the sets 103,104 and the valve V of the means 101. The controller 105 implements the control by means of a control algorithm stored in its memory means, under a system program controlling the device. The control algorithm is formed to provide the desired control. For the con- trol, the controller 105 receives, when necessary, one or more set sig- nals 107 and 108 from the outside, on the basis of which the valves of the sets are opened and closed, or the valve of the means 101 is shifted, when necessary. If necessary, the controller 105 is also used to automatically adjust the rate at which the volume flow of the connec- tions in the valve system 100 is changed, that is, to adjust the accel- eration and deceleration ramps of the actuator. The asymmetry of the transmission of the volume flows of the sets is also adjusted in each situation; for example, the transmission of the set 104 on the return side is guided to a higher level than the transmission of the set 103 on the pressure side, to reduce pressure losses. Alternatively, to avoid cavitation, the transmission of the set 103 is guided to be higher than that of the set 104. Furthermore, the control signal 107,108 to be input to the controller 105, such as a set signal (for example, a voltage signal from-10 V to +10 V or a current signal), is applied in a control algo- rithm to compute the level shown in Fig. 2 and the direction of motion of the actuator, followed by the control of the valves of the sets 103 and 104 and the means 101. The control of the valves V, V1-V4 may be implemented, for example, by relays of the control means 105, through which the operating voltage is input to the valves.

Parameters related to the computing of the control algorithm of said controller 105, such as a deceleration ramp, an acceleration ramp, the asymmetry, the size of the control zero area or dead area, the ratio between the set signal and the transmission capacity, the valve combi-

nation at different transmission levels, the separation of the control of the sets, and the special couplings shown in Figs. 4a to 4d, are pref- erably input to the controller 105, for example, by means of a pro- gramming device or a control system 109, such as a computer system, which forms the control system of the upper level. At least some of the parameters can be set by means of control screws or control buttons connected directly to the controller 105. It may even have a display and a keypad for manual input of the parameter values. Said control sys- tem 109 can be used simultaneously to control also other valve sys- tems or devices to which the valve system 100 is connected. To set the valve system 100 in different states, for example to create various con- nections, the set value and the parameters can be input to the control- ler 105 in digital format, for example through a bus 107 or 108, but in some embodiments, also in analog formats by a set signal 107 or 108.

Each parameter may also be provided with a separate signal line.

It is also possible to use agreed default values for the parameters, wherein the control is performed simply by one set value (for example, a voltage signal from-10 V to +10 V or a current signal), and the con- troller takes care of the more precise control of the system 100 in an optimal way by means of the control algorithm. The set value may also be entered under manual control, for example by means of a control stick controlled by the operator to achieve the desired rate. The desired rate is thus, for example, slow or fast, depending on the position of the control stick.

In one embodiment of the invention, the valve system 100 consists of separate valves V and V1-V4 which are preferably fixed to a joint base plate which is also provided with the necessary drilling and channel- lings to form said connections. The system becomes a uniform and compact unit. Preferably, the base plate is also provided with some of the ports (P, T, A, B) of the valve system 100, for example for a tube fastening or a bed plate fastening. It is obvious that the necessary con- nections may also be formed by means of tubes and hoses by joining the valves. It is obvious that the valve system may also be provided with other ports to form special connection in addition to the presented connections 1 to 4. Thus, particularly the possibilities to couple the

valve V of the coupling means 101 will be increased. For example, the valve V could be used to close the ports A and B and to set up a con- nection between the ports P and T via said sets 103 and 104.

Calculation of the controls of the valves in an example system Figure 5 shows a typical circuit applying an on/off valve system for the control of a cylinder 200. For simplicity, the tank pressure used in the calculation is the value 0 bar, and the pressure losses of the valve 101 are assumed to be insignificant, and the volume flows Q selected for the valves (V1, V2, V3 and V4) are 0.5,1,2 and 4 I/min, respectively, when the pressure difference over the valve is 5 bar. Alternatively, the series may be, for example, 2,4,8 and 16 I/min. The proportion of the tank pressure and the pressure losses may also be considered a sepa- rate factor. Next, we shall discuss the modelling of the system, wherein the following denotations will be used: AA 1) Cylinder area on the piston side [m2] AB 1) Cylinder area on the piston rod side [m2] KVP 1) Volume flow coefficient of the smallest valve in the valve set on the pressure side [m3/ (s#Pa0.5)] KVT 1) Volume flow coefficient of the smallest valve in the valve set on the return side [m3/ (s#Pa0.5)] nP 1) Number of valves on the pressure side nT 1) Number of valves on the return side avp 2) Opening of the valve set on the pressure side (inte- ger from 0 to 2(nP-1)) avT 2) Opening of the valve set on the return side (integer from 0 to 2 (n ; 1)) y Area ratio of the cylinder AA/AB z Auxiliary variable avp. (aVT#KVT) v 3) Cylinder rate [m/s] PA 3) Pressure in the cylinder chamber on the piston side [Pa] PB 3) Pressure in the chamber on the piston rod side [Pa] ps 1), 2) Feeding pressure [Pa] Qp Volume flow through the valve set on the pressure

side [m3/s] QT Volume flow through the valve set on the return side [m3ls] F 1), 2) Net force effective on the cylinder (external forces and frictions) When the piston 201 moves in the positive direction, the volume flows willbe: The force equation being: (3) F = AAPA-ABPB In the balanced state of the circuit (positive direction), the pressures and the rate are: When the piston 201 moves in the negative direction, in turn, the vol- ume flows will be: p#ps - pB = -vAB and (8) QT = avTKVT#pA = -vAA The force equation being: (9) F = AAPA-ABpB In the balanced state of the circuit (negative direction), the pressures and the rate are:

From the preceding formulas, it will be seen that in the balanced state, the pressures depend on the opening of the pressure side (P) and the return side (T), the feeding pressure, and the force effective on the cylinder 200, wherein the following functions can be marked: v = v(avP,avT,pS,F) (13)pA = pA(avP,avT,pS,F) pB = pB(avP,avT,pS,F) The required computation is implemented in the control algorithm of the control means 105, in which the invariable parameters describing the system (marked above with the reference 1)) are input. The computa- tion is implemented with different combinations of the variable parameters (marked above with the reference 2)), which results in the outputs (marked above with the reference 3)) After optimization and the selection of the combination, the control means 105 generate control signals 106 to guide the valve means 102, ie. the valves, to the desired position. When the desired set value is changed, the opera- tions are performed again. The set value can be continuously changed and monitored by the control algorithm. In the following, the procedure of the control algorithm will also be presented to achieve the desired set value.

The computation of the combinations can also be performed sepa- rately, within or outside the control system 109. The computation is made, for example, with a computer device, and the control means 105 are used to store, for example in a table, the computed rate values and the values of said parameters which were used for the computation (the pressures, the opening combination used, the forces, etc.). The values of the table can be placed in an optimation algorithm to find out the order of supremacy of the opening combinations. The desired combination is selected for use, and on the basis of the table, the valves are guided to the corresponding position. The arrangement reduces and accelerates the computation of the control means 105.

The control means 105 may also comprise a unit which is easy and ready to use, including the calculation of the opening combinations, wherein the data about the actuator and valve set to be used is first

input therein. Thus, the control means 105 may already comprise sev- eral alternative models on the pressure medium circuits 100 to be used, or they can be defined for example by means of parameters or by modelling. The control means 105 take care of the computation and storage of the necessary computation algorithms.

Control of the rate in the example system We shall discuss the operation in the positive direction. The negative direction can also be processed in a similar way by taking it into account in the above-mentioned formulas. If it is desired to control the rate without feedback, the feeding pressure ps and the force F effective on the cylinder 200 should be known or measured. According to the invention, the above formulas can be used to calculate the pressures of the balanced state and the rate for all the possible opening combina- tions of the on/off valves on the P and T sides. Of these values, the best possible one can be selected by using a desired criterion.

According to the invention, a desired penalty function J is used here, which is for example a function on the pressures and rate of the bal- anced state, and possibly also on the change in the pressure, wherein: (14) J = J (v, pA, ps) Figure 6 shows the speed v of the cylinder 200 of Fig. 5 with different combinations of openings of the valve set 103 on the P side and the valve set 104 on the T side, wherein a large number of separate, dis- continuous or discrete values are obtained, because of the on/off valves. By arranging the combinations in an order according to the rate to be obtained with the actuator, the stepped chart of Fig. 7 is obtained, in which each step represents a given combination. A particular advantage of the invention is that substantially the same rate is repre- sented by several different combinations whose pressure levels may be very different and of which it is then possible to select the most suit- able one, as long as for example the rate error is within allowed limits.

Pressure variations on the piston side in the example system are illus- trated in Fig. 8. As a selection criterion, for example a minimum change in the pressure is used, compared with the prevailing pressure level.

Also other criteria can be used in the selection of the combinations to be used next.

In the way of controlling the P and T sides separately, one of the two valve sets is normally fully open. Thus, only the step of Fig. 2 is in use, wherein Vmax thus corresponds to the outermost combination of the T or P valve set, which is parallel to either the T-axis or the P-axis. The present invention is not limited to these steps, but all the steps will be taken into account. It is also possible to move on the same rate level by means of different combinations, because equal columns denote vari- ous combinations (and also different pressure levels) but substantially the same rate. By the control method of the invention, it is possible to advance from one column to another, simultaneously optimizing the control.

The selection of the opening combination to be used for the control each time is a difficult task, to which the penalty function J of the pre- sent invention provides a solution. On the basis of this, it is possible to select that opening combination to be coupled next, which will imple- ment the control of the cylinder rate in the desired way. The following discussion will also apply analogically to hydraulic engines, which application will be obvious for a man skilled in the art on the basis of what has been presented above.

In the first embodiment of the invention, only the rate error is taken into account in the penalty function J, for the control of the rate: (15) J(v) = K0#(#vref'-v#)x, wherein zizis the desired cylinder rate. The variable X is the desired exponent, for example 2, and the coefficient Ko is a weight coefficient.

To compute the rate, one must also know the force F effective on the cylinder. Next, we shall compute the pressures pA, PB and the rate v of the balanced state with all the possible opening combinations. After this, these different values (PA, PB, v) are used to compute the value of the penalty function J. Next, we shall select that combination of the

valve sets (for example, 103 and 104) which best meets the criteria of the penalty function J. In this case, that opening combination will be selected, for which the function J has the lowest value, wherein the rate error is the smallest. It will be obvious that a maximum value can be set for the rate error, wherein it is possible to select any suitable opening combination, for which the function J is smaller than it. The selection of the suitable opening combination may also be based on other criteria.

In another advantageous embodiment of the invention, not only the rate control but also the cavitation is taken into account, wherein those opening combinations are rejected in which the pressure pA and/or the pressure ps is closer to zero pressure than the set value pmin. In this case, the penalty function J is, for example : The value set for the penalty function J is a very large number (or this alternative can be excluded from the further processing), if there is a risk of cavitation with this opening combination, wherein this combina- tion is clearly not included in those to be selected.

In a third embodiment, target pressures PA, ref and pgsref are set for the pressures PA and PB, not to deviate from them more than desired. The magnitude of the deviation will depend on the coefficients of the pen- alty function. Among other things, the target pressures are used to influence energy savings, wherein the pressure should be kept as low as possible, but by avoiding cavitation. The aim at the target pressure can be combined with the need to keep the rate approximately within a desired range and represented with the formula (without conditions for preventing cavitation, which may be added in the formula) : 17) J = K0#(#vref'-v#)x + K1#(#pA,ref - pA#)Y + K2#(#pB,ref - pB#)Z,

in which the parameters K1 and K2 (weight coefficient) are used to determine, how aggressively the pressure is to be kept within a desired range. When the pressure level is not at issue, the coefficient has the value zero.

The on/off valves operate in a stepped manner, and with different com- binations it is possible to select a set of discrete steps, wherein great variations in the pressure level, as a counter-balance for good rate control, is not always acceptable. In a fourth embodiment of the inven- tion, changes in the pressure level, ie. pressure jumps, are reduced by adding the change in the pressure level in the penalty function J as follows : (18) J = K0#(#vref'-v#)X + K3#{K4(#pA - pA,ed#)Y + K5(#pB - pB,ed#)z}, in which PA, ed and PB, ed are the pressures selected in the previous computation round, that is, the discrete step used by the control, from which the aim is to move on to the next one to implement cylinder con- trol and a change in the rate. For example, the penalty function may be supplemented with the optimization of the target pressures, wherein the variation takes place in the range of a given pressure value.

For energy saving, the maximum of the feeding pressure ps should be kept as low as possible. The aim is achieved by requiring that the low- est pressure in the system must be relatively low and constant. Nor- mally, the lowest pressure is on the return side, when the force is resistant. Thus, according to one example, the aim is to keep PB as desired (pB, ref) within the set limits (pB, ref-pB), and pA is allowed to vary freely.

Rate control in the example system Next, we shall discuss different ways of determining the rate vref' required in the selection of the opening combinations. The simples case is open control, in which the desired rate vref'is obtained directly from the control means 105,109 or from the operator, for example by

means of a control stick. The rate error is the smaller, the better the loading force F and the feeding pressure ps are known. Thus, by open control, precise rate control will only be achieved, if the loading force is accurately known or measured. On the other hand, for example in manual control, very precise rate control will not be necessary.

In the rate control, the force effective on the cylinder is not always known. A more precise rate control will be achieved by measuring the realized rate and by selecting valve combinations by minimizing the rate error. If the target rate vref is higher than the measured rate v, one should select an opening combination which provides a higher rate.

The problem, however, is the loading force F, whose value may be poorly known. In the invention, however, we have noticed the feature which is essential for the closed control that the order of the opening combinations is independent of the loading force, wherein in the for- mulas (4) to (6) and (10) to (12) it is possible to use a rough estimate for the loading force, or to set the value of the loading force to zero in the calculation. This makes it possible to determine and select in advance the opening combinations to be used, by arranging them for example in such a way that the variations of the rate v are small and/or that cavitation is prevented. These combinations are arranged in an order, and, for example, to increase the rate, the opening combinations are selected in this order towards a higher rate, until the desired rate is realized. The same principle is also applied for deceleration. At the start of the control, it is possible to select, for example, every fifth value and to gradually make the selection more accurate, wherein, in the end, the next value will always be selected.

If the force F is known in advance or can be estimated in a reliable way, the sets of the different opening combinations can be predeter- mined (penalty function 4. By means of the sets, one achieves for example a good rate control (formula (15)) and optimal pressures (for- mula (16)), when the loading force F is positive, negative or zero, each as a set of its own. On the basis of the estimated loading force, it is possible to specify and always select the most suitable set to be used in the open or closed control. Thus, the suitable opening combination can be selected directly without advancing in an order, naturally

depending on the desired rate profile as a function of time. At the same time, it is possible to optimize for example the pressures.

In closed rate control, the rate vref'required for the selection of the opening combinations is generated by a rate controller, whose input is the rate error (vref-v) and whose control algorithm is aimed at mini- mizing the rate error. The function used for the computing algorithm is of the form vref'= fV (vref-v), in which fv is a static or dynamic descrip- tion. The control precision can be improved further by using feed- forward, wherein the desired rate vref is added in the output of the rate controller : vref'= vref + fv (vref-v). The advantage in this way of control- ling is that the rate controller operates with low difference variables.

The desired rate may vary as a function of time so that for example the desired rate ramp vre, +f) could be followed. The control means 105 may be arranged to continuously determine this desired rate. The desired rate as a function of time v, e (t) can also be stored, for example, as a function or a table, to be read by the control means. This desired rate value, whether constant or variable, may also be a control signal 107 and/or 108, which is generated with the same or a separate control system 109.

Position control in the example system In the control of actuators, movements are often used, in which the position is changed according to a path or a curve, for example position and rate ramps. Based on the rate control in the system, one also achieves good position control, even though the control is formed of discrete combinations. The system is thus particularly suitable for position control, when the actuator is continuously in motion, following a desired position.

At each moment of time, the desired rate vref can be generated on the basis of the desired position path great). The desired path xref (t) can also be stored, for example, as a function or a table, to be read by the control means 105. For the feedback, the system also comprises

means for determining the position of the actuator, the measured posi- tion x being, in turn, compared with the desired position xref.

In one embodiment of the invention, the position referencexref (t) is used to determine the required rate, which is based, for example, on the fact that the actuator must have a given rate to reach the next posi- tion within a given time. Also, the path can often be presented as a curve which is, for example, a position ramp, a rate ramp, or a polyno- mial of the third or fifth order, and which can, in most cases, also be derived 1 to 3 times. The derivation will result in the rate curve to be used, indicating the target rate vref for each moment of time. The rate profile can also be stored to be ready in for example a table, or it is determined by computation in the control means 105, in which the position path is input.

In the closed position control, the operation is the same as described above in connection with the rate control, but the rate controller is replaced by the position controller in the form vref'= fx (xref-x). Using the feedforward, the form is: vref'= vref + fx (xref-x). One simple position controller is, for example, of the form: fx = Kp (xref-x), in which Kp is the amplification.

Control of the example system by using pressure measurement By measuring the pressures PA, pB, the loading force can be deter- mined, for example, by the formula (3). The result is to achieve good rate control without feedback from the rate. The optimization of the pressures is also more successful, because measured information is also obtained about the pressures. It is also possible to use lower counterpressures without the risk of cavitation, because the pressures have been measured. Alternatively, it is possible to use a loadcell in a way known as such for measuring the force.

In the control according to the invention, advantages are achieved especially in the energy saving; in particular, the feeding pressure ps can be reduced. When using a pump with a fixed volume and setting the feeding pressure ps with a shut-off valve, it is possible to use a

lower feeding pressure than with symmetrical valves of prior art, in which the pressures on both the pressure and return sides are affected by pressure losses over the shared slide. The separate valve sets can be controlled individually, wherein particularly the counterpressure can be minimized, still avoiding cavitation.

In the use of an adjustable-displacement pump, the possibilities for energy saving are greater, because the operating pressure of the actuator can be optimized to be as low as possible, to be slightly exceeded by the feeding pressure of the pump.

Furthermore, it will be obvious for anyone skilled in the art that although the invention has been illustrated in the above description by means of an advantageous valve system, the invention can be applied within the scope of the claims. Furthermore, it should be noted that the above-presented penalty function is only an example of the mathematical format in which the penalty function can be brought for the computation. With the mathematical form varying, it is, for example, possible that the most optimal result is achieved by maximizing the formed function. The penalty function and its coefficients may also depend, for example, on the direction of motion and the rate.