Yaodong Pan

Stability and performance improvement of extremum seeking control with sliding mode

Extremum seeking control is an adaptive control strategy for a system to iteratively extremize a function in real-time. In this paper we investigate stability and performance improvement by using sliding mode concepts in extremum seeking controllers.

Variable-Structure Control of Electronic Throttle Valve

In recent years, the use of electronic throttle valve systems has been very popular in the automotive industry. However, there exist difficulties in controlling the electronic throttle valve because of multiple nonsmooth nonlinearities including stick-slip friction, backlash, and a discontinuous nonlinear spring involved in the system. To alleviate the aforementioned difficulties in controlling the opening angle of a throttle plate and to realize a highly robust controller against uncertainties in the throttle bodys mathematical model with limited cost, the variable structure concept is utilized after the employment of feedback backstepping techniques in the intermediate stages of design. Furthermore, the sliding mode observer design technique with equivalent control is exploited in order to estimate the values of necessary states that are not measured. The performance of the proposed controller is evaluated by performing some experiments on the throttle valve setup. First, small scale reference signals that pass through the nonsmooth nonlinearities are tracked even in different disturbed situations by the throttle valve plate, and then, the performance of the controller is tested for large-scale reference signals to observe the reaction time of the controller to sudden changes. For all the cases, the controller works rather well and meets the performance specifications.

Discrete-time sliding mode control of electronic throttle valve

Unlike a conventional throttle valve which is connected to the accelerator directly by a wire, an electronic throttle valve is driven by a DC motor and can move independently from the position of the accelerator. This feature provides flexibility in determining the amount of air flow into the engine. However, it is difficult to control the electronic throttle valve as the valve system involves nonsmooth nonlinearities such as stick-slip friction, gear backlash, and nonlinear spring. The electronic throttle valve system is modeled with the nonsmooth nonlinearities considered. A discrete-time sliding mode controller and observer are designed to realize robust tracking control of the valve system. Experimental results are presented.

A Control Authority Transition System for Collision and Accident Avoidance

Summary This paper addresses the problems ensuing with the human being who is controlling the technical system, especially problems related to his skill levels, driving habits, capabilities and decisions (especially when impaired by drugs, fatigue or physical handicaps). It may be possible to improve the ability of a driver to operate a vehicle safely if certain parameters in the control of the vehicle are adjusted according to the driver’s normal characteristics. Value has already been established for tailoring certain attributes such as seat, pedals, steering wheel position and mirrors to a given driver through memory functions. This research concentrates on assessing a driver’s operating characteristics and modifying the control to improve safe operation of the vehicle on a real-time basis.

Sliding mode control of electronic throttle valve

This paper focuses on the design of a controller for the throttle valve that is used in combustion engines for the purpose of adjusting air-fuel ratio. Due to the right-hand side discontinuities and uncertainties in the throttle bodys mathematical model, sliding mode concept is utilized after the employment of feedback linearization techniques in the intermediate stages of design. Finally, the performance of the proposed controller is evaluated with the simulation and experimental studies.

Nash solution by extremum seeking control approach

In this paper, we propose an algorithm to solve the Nash equilibrium solution for an n-person noncooperative dynamic game by the extremum seeking control approach with sliding mode. For each player, a switching function is defined as the difference between the players cost function and a reference signal. The extremum seeking controller for each player is designed so that the system converges to a sliding boundary layer defined in the vicinity of a sliding mode corresponding to the switching function and inside the boundary layer, the cost function tracks the reference signal and converges it to the Nash equilibrium solution.

Discrete-time extremum seeking algorithms

In this paper, we propose three kinds of discrete-time extremum seeking control algorithms. As the static map from the setpoint to the cost function can be approximately represented by a polynomial, the extremum can be easily found according to the polynomial. In the first algorithm, the extremum seeking control is realized by the least square estimation method. In the second and third algorithms, the extremum point is estimated by approximating the static map as a parabola and an ellipse, respectively. The simulation results show the feasibility of the proposed algorithms.

Non-standard safety technology

This paper explains the nonstandard safety technology which assists drivers to enhance drive safely when the driver is physiologically impaired, for example, drunk, fatigued, irritable, asleep, etc. When a driver operates a vehicle with some actions which may not result in an accident immediately but may reveal that the driver is drunk, fatigued, or asleep, the non-standard safety enhancement system makes a decision from the actions to determine whether the driver is physiologically impaired based on the available information of vehicle, road, and others before the driver causes an accident.

A control authority transition system for collision avoidance

Drivers differ significantly in their skill levels, driving habits, capabilities and decisions especially when impaired by drugs, fatigue or physical handicaps. Late or inadequate responses to hazards and dangers may cause severe results. It may be possible to improve the ability of a driver to operate a vehicle safely if certain parameters in the control of the vehicle axe adjusted according to the drivers normal characteristics. Value has already been established for tailoring certain attributes such as seat, pedals, steering wheel position and mirrors to a given driver through memory functions. This research concentrates on assessing a drivers operating characteristics and modifying the control of the vehicle to improve safe operation of the vehicle on a real-time basis.

Variable structure control of electronic throttle valve

The use of an electronic controlled throttle valve can increase the efficiency and performance of an engine since the control of the throttle valve, which is determined according to the desired amount of air flow calculated from the desired speed and engine torque, is independent from the position of the accelerator. In this paper, a variable structure (VS) controller is designed for the electronic throttle valve. The model of the electronic throttle valve is discussed at first. Then the VS controller is proposed by using the backstepping approach. Finally, simulation and experimental results are given.