The State Estimation for Electric Stability Program Using Kalman Filtering

In electric stability program (ESP), more vehicle information is demanded, while some signals can not be measured directly with sensors due to some physical reasons or the cost reason of the sensors. Then some other ways have to be found to get such signals. The Kalman filtering theory can provide such a method: some state variables of a system can be estimated with the information of some related measured variables. So, this method can be used to solve the problem of estimating the variables of the vehicle. As the vehicle system is highly non-linear, the Kalman filtering method of the non-linear system should be used: it linearizes the non-linear system at each step of the arithmetic flow. In this paper, a state estimator have been constructed, using the Kalman filtering method of the non- linear system, with a set of vehicle motion equation and measurement equation, for which the measurements are pointed out in advance. Then, using the software of MATLAB, a fourteen-degree of freedom vehicle model is used to emulate the true vehicle and noises are added to simulate the true sensor noises. The constructed estimator is programmed with M language in MATLAB. Then, the measurements are gained directly from the simulating vehicle model and used as the inputs of the estimator. The estimator proves well after the compare of the simulation results: the true values of the variables, which are obtained from the signal of the simulating vehicle model, and the estimated values from the estimator are equal respectively with acceptable errors. No doubt, such a way to get the state variable signals of vehicle can solve the problem for ESP and may make for some other advanced vehicle control. Besides, if the method is used aiming at the signals of some costly sensors, such sensors can be instead, which will cut the cost of the advanced control systems.