Controller, control system and method for vehicle control

1. A control system for a vehicle, the control system comprising:

a first controller configured to generate, by solving a first optimization problem, an optimal trajectory of the vehicle along a path;

a second controller configured to, based on the optimal trajectory generated by the first controller, generate motoring and braking commands to a motoring and braking system of the vehicle for controlling the vehicle to travel along the path; and

an auto-tuner comprising a processor configured to solve a real-time second optimization problem to determine at least one parameter of the first controller.

2. The control system of claim 1 , wherein

the second controller is a controller separate from the first controller.

3. The control system of claim 1 , wherein

the optimal trajectory comprises an optimal profile of at least one of speed, acceleration or jerk of the vehicle along the path.

4. The control system of claim 1 , wherein

the second controller is an optimal adaptive controller configured to, based on the optimal trajectory generated by the first controller, solve a third optimization problem to achieve at least one control objective under at least one constraint, to generate optimal motoring and braking commands as the motoring and braking commands for controlling the vehicle to travel along the path.

5. The control system of claim 1 , wherein

the second controller comprises a neural network machine learning controller.

6. The control system of claim 1 , wherein

the second controller comprises a Proportional-Integral-Derivative (PID) controller.

7. The control system of claim 1 , wherein

the second controller is configured to, in a self-training mode, generate randomized or predetermined motoring and braking commands to the motoring and braking system, for estimating at least one parameter of at least one of

a first model of motion dynamics of the vehicle, or

a second model of the motoring and braking system of the vehicle.

8. The control system of claim 7 , wherein

the processor of the auto-tuner is configured to, in the self-training mode, estimate the at least one parameter of the at least one of the first model or the second model based on

the randomized or predetermined motoring and braking commands, and

sensor data indicative of responses of the vehicle or the motoring and braking system to the randomized or predetermined motoring and braking commands.

9. The control system of claim 8 , wherein

the processor of the auto-tuner is configured to provide at least one of

the first model to the first controller to enable the first controller to generate the optimal trajectory of the vehicle along the path, or

the second model to the second controller to enable the second controller to, based on the optimal trajectory generated by the first controller, generate the motoring and braking commands to the motoring and braking system of the vehicle for controlling the vehicle to travel along the path.

10. The control system of claim 9 , wherein

the processor of the auto-tuner is configured to, while the vehicle is traveling along the path, repeatedly solve the real-time second optimization problem to optimize at least one parameter of at least one of the first model in the first controller or the second model in the second controller.

11. A controller for a vehicle, the controller comprising:

at least one processor configured to

using a first model and by solving a first optimization problem, generate an optimal trajectory of the vehicle along a path,

based on the optimal trajectory and using a second model, generate motoring and braking commands to a motoring and braking system of the vehicle for controlling the vehicle to travel along the path, and

while the vehicle is traveling along the path, repeatedly optimize, by solving a second optimization problem, at least one parameter of at least one of the first model or the second model,

wherein the at least one processor is further configured to, in a self-training mode, estimate the at least one parameter of the at least one of the first model or the second model based on

randomized motoring and braking commands to the motoring and braking system, and

sensor data indicative of responses of the vehicle or the motoring and braking system to the randomized motoring and braking commands, and

wherein the vehicle is a railway vehicle.

12. The controller of claim 11 , wherein

the first model comprises a model of motion dynamics of the vehicle, and

the second model comprises a model of the motoring and braking system of the vehicle.

13. The controller of claim 12 , wherein

the model of the motoring and braking system comprises at least one of

a model of a motoring system of the vehicle,

a model of a dynamic braking system of the vehicle,

a model of a friction braking system of the vehicle,

a model of a service braking system of the vehicle,

a model of a regenerative braking system of the vehicle, or

a model of an emergency braking system of the vehicle.

14. The controller of claim 11 , wherein

the at least one processor is further configured to optimize the at least one parameter under at least one constraint selected from the group consisting of:

a maximum error between a simulated drive of the vehicle in accordance with the first model and an actual drive of the vehicle,

a maximum error between a simulated operation of the motoring and braking system in accordance with the second model and an actual operation of the motoring and braking system, and

maximum and minimum values of the at least one parameter.

15. The controller of claim 11 , wherein

the at least one processor is configured to

solve the first optimization problem to achieve at least one first control objective under at least one first constraint, and

solve the second optimization problem to achieve at least one second control objective under at least one second constraint.

16. A method of controlling a vehicle, the method comprising:

generating, by a first controller solving a first optimization problem, an optimal trajectory of the vehicle along a path;

generating, by a second controller and based on the optimal trajectory, motoring and braking commands to a motoring and braking system of the vehicle for controlling the vehicle to travel along the path; and

while the vehicle is traveling along a path, optimizing at least one of the first controller or the second controller in real-time, said optimizing comprises solving a second optimization problem,

wherein the vehicle is a railway vehicle.

17. The method of claim 16 , wherein

said generating the optimal trajectory of the vehicle is executed in accordance with a model of motion dynamics of the vehicle, and

said generating the motoring and braking commands based on the optimal trajectory is executed in accordance with a model of the motoring and braking system.

18. The method of claim 17 , wherein

said optimizing comprises:

optimizing a parameter of at least one model of the model of motion dynamics of the vehicle or the model of the motoring and braking system, and

updating, using the optimized parameter, the at least one model included in at least one corresponding controller of the first controller or the second controller.

19. The method of claim 17 , wherein

said optimizing comprises:

optimizing at least one model of the model of motion dynamics of the vehicle or the model of the motoring and braking system, and

training, using a neural network and based on the optimized at least one model, at least one corresponding controller of the first controller or the second controller.

20. The method of claim 17 , wherein

said optimizing comprises:

performing a non-linear optimization of the model of the motoring and braking system, and

tuning the second controller, which is a Proportional-Integral-Derivative (PID) controller, in accordance with a simulation based on the optimized model of the motoring and braking system.