This invention relates to a method for use in controlling the operation of an autonomous vehicle. In particular, the invention relates to a control method whereby an autonomous vehicle carrying a sensor sensitive to an environmental parameter can be moved to identify the location of, for example, a maximum or minimum value of the environmental parameter.

Autonomous vehicles are well known. Certain designs of vehicle are intended for use on land, for example having wheels, tracks or the like, and a drive arrangement whereby the wheels, tracks or the like can be driven to allow the vehicle to be propelled over the ground, the control being such that steering of the vehicle can be achieved. Other known autonomous vehicles are intended to be used in, for example, bodies of water, the vehicles being capable of being controlled in such a fashion as to adjust their position in two or three dimensions.

Such vehicles may be provided with one or more sensors whereby environmental parameter measurements can be made indicative of the environmental parameter values at the location of the vehicle. By moving the vehicle, mapping of the parameter values may be achieved. By way of example, if a body of water is suffering from pollution, a sensor sensitive to the concentration of the pollutant may be carried by the vehicle, and the vehicle moved within the body of water to map the concentration of the pollutant. In an alternative application, the parameter to be measured may be altitude, and the vehicle may be driven over a ground surface to map the shape of the ground surface. It will be appreciated that these are merely two examples of applications in which an autonomous vehicle may be used. Other applications in which autonomous vehicles may be used include, for example, exploring the spread of nuclear radiation, the expansion of algae plumes and the spread of volcanic ash. It will be appreciated that this is not an exhaustive list of the applications in which such vehicles may be employed.

A number of techniques for controlling the operation of such autonomous vehicles are known. One control technique that is known involves the use of a sliding mode based control algorithm. By way of example, P.P. Menon et al "Boundary Tracking using a Suboptimal Sliding Mode Algorithm", 53 ^rd IEEE Conference on Decision and Control, Los Angeles 2014, pp5518-5523, describes a control technique whereby an autonomous vehicle is controlled in such a manner as to identify the shapes and locations of contours of a known parameter value, for example a known pollutant concentration. The technique involves the use of a sliding mode based algorithm. In such an arrangement, a predetermined concentration level is set, and the algorithm is used to control the movement of the vehicle in such a manner as to follow the predetermined concentration. By repeating the procedure with a series of different parameter values, a series of concentration contours or boundaries may be derived. Other control methods are known to achieve similar ends. These methods typically involve estimating a gradient at which it is expected that the pollutant concentration will increase as the vehicle is moved towards the source. The movement of the vehicle is then controlled in such a fashion that the gradient of the concentration of the pollutant in the fluid through which the vehicle is moved is maintained at substantially the estimated gradient. A technique of this general type is described in greater detail in R. Bachmayer and N.E. Leonard "Vehicle Networks for Gradient Descent in a Sampled Environment", Proceedings of the 41 ^st IEEE Conference on Decision and Control", Las Vegas 2002, pp112-1 17. Whilst this type of method can be used to identify the location of the maximum or minimum sensed parameter value, it can only be used where the gradient can be estimated. If the gradient is not known then controlling a vehicle using this method is not possible.

It is an object of the invention to provide a control method for use in controlling the operation of an autonomous vehicle, whereby the location of a maximum or minimum parameter value can be identified using the vehicle.

According to the present invention there is provided a control method for use in controlling the operation of an autonomous vehicle carrying a sensor sensitive to a parameter associated with the environment in which the vehicle is located, the method comprising the steps of using a sliding mode control algorithm to control the direction in which the vehicle is moved, wherein the sliding variable of the sliding mode control algorithm is the difference between the sensed parameter value at the location of the vehicle and a time varying reference trajectory in the form of a monotonically changing function. By way of example, the sliding mode based control algorithm may take substantially the form described in P.P. Menon et al "Boundary Tracking using a Suboptimal Sliding Mode Algorithm", 53rd IEEE Conference on Decision and Control, Los Angeles 2014, pp5518-5523. However, rather than having a fixed reference trajectory, the use of a time varying reference trajectory in the form of a monotonically changing function results in the vehicle gradually moving towards the maximum or minimum value of the sensed parameter rather than following a fixed contour or boundary.

The invention will further be described, by way of example, with reference to the accompanying drawing, Figure 1 , which is a diagram illustrating the path followed by a vehicle using the method of an embodiment of the invention.

As described hereinbefore, the present invention relates to a method for use in controlling the operation of an autonomous vehicle. By way of example, the vehicle may comprise a marine vehicle capable of being self-propelled through water by appropriate controlled propellers. The vehicle includes a steering mechanism to allow the direction in which the vehicle moves to be controlled. By way of example, the vehicle may include two or more propellers, the relative speeds of which may be modified to turn the vehicle in a desired direction. Alternatively, for example, rudders or the like may be provided. In this example, it is assumed that the vehicle will be driven at a substantially uniform speed. However, this need not always be the case.

The vehicle carries a sensor sensitive to a parameter associated with the environment in which the vehicle is located. By way of example, the sensor may be operable to sense the concentration of a pollutant contained within the marine environment in which the vehicle is located. It will be appreciated, however, that this represents merely one example of an application in which the invention may be employed, and that the sensor may take a range of other forms. In accordance with the invention, the vehicle is controlled in such a manner as to allow it to move towards, and hence identify the location of, a point of maximum pollution concentration. This is achieved through the use of a sliding mode based control algorithm, for example substantially in the form described in P.P. Menon et al "Boundary Tracking using a Suboptimal Sliding Mode Algorithm", 53rd IEEE Conference on Decision and Control, Los Angeles 2014, pp5518-5523. However, in that paper, the algorithm is used to identify boundaries of uniform concentration rather than to identify the location of a point of highest concentration. Accordingly, the method is modified in such a manner that the sliding variable of the sliding mode control algorithm is related to a reference trajectory in the form of a monotonically changing function, rather than taking a fixed value. Specifically, it is the difference between the sensed parameter value at the location of the vehicle and the time varying reference trajectory.

By using a sliding variable of this form in the control algorithm, rather than the vehicle being steered or controlled by the algorithm to follow a path of substantially constant concentration, it is steered or controlled in such a fashion as to result in the concentration of the pollutant in the environment through which the vehicle is passing substantially constantly increasing. Figure 1 illustrates the path 10 followed by the vehicle. It is apparent that the path 10 does not follow any particular boundary or contour 12 of substantially uniform concentration, but rather is constantly crossing from one contour 12 to another of higher concentration until a point is reached at which it is unable to identify a higher concentration contour and so follows a looped path at or adjacent the point 14 of maximum concentration. As set out hereinbefore, in order to achieve source seeking, the tracked value Y _re f(t) is monotonically increased from the value at the vehicle's initial deployment Y _re f(0)= γ(χ(0), y(0)), where (x(0), y(0)) is the initial deployment of the vehicle and γ(χ, y) is the concentration at the point (x, y). Provided a sliding motion occurs, the tracked value Yref(t) constantly increases or stays the same. A reduction in the tracked value is not permitted Once the maximum value γ is exceeded a sliding motion can no longer be sustained, and so sliding will be broken (and never regained). At this point the value of Yref(t) remains frozen. As sliding is broken, the vehicle will circle around the value (or close to) the maximum value γ. The rate at which the tracked value Y _re f(t) increases can be selected by the operator. As a general rule, the rate of increase should be directly proportional to the vehicle's velocity and inversely proportional to the difference between γ _Γβ ί(0) and the maximum value γ. Having a high rate of increase will result in a faster detection of the approximate location of the maximum, but the accuracy with which the location is identified will be poorer. Also, the diameter of the path followed around the maximum being increased. It will be appreciated, therefore, that a trade-off is involved in the setting of the rate of increase between achieving a result rapidly and achieving a result with a good degree of accuracy. If the approximate location of the maximum concentration is known, then the methodology may be modified to enhance the speed with which the location is identified. By way of example, the rate of increase may, initially, be set at a high value until the vehicle reaches substantially the location in which it is known that the maximum is to be found, the rate of increase subsequently being reduced to allow accurate location of the maximum.

Whilst the description hereinbefore is of an arrangement in which the location of a maximum concentration of a pollutant is to be found, it will be appreciated that the invention is suitable for use in a number of other applications. In some circumstances, it may be arranged to identify the location of a minimum rather than a maximum parameter value. Furthermore, the range of parameters with which the invention may be employed is large. By way of example, it may be used to identify peak pollutant concentrations, sources of radiation, points of highest altitude, points of highest pressure, points of greatest magnetic or electrical field strength or a range of other parameters. Whilst the invention is easy to envisage in a two dimensional space, it will be appreciated that the methodology of the invention may be modified to allow the location of a maximum or minimum to be identified in a three dimensional space, if desired. Whilst the description hereinbefore makes reference to a marine type autonomous vehicle, it will be understood that the invention is also applicable to vehicles intended for use on land. The invention could also be applied to drones or other airborne vehicles. Although specific arrangements falling within the scope of the invention have been described herein, it will be appreciated that a wide range of modifications and alterations may be made without departing from the scope of the invention as defined by the appended claims.