Parametric Brake Optimization Based On Ant Colony Algorithm

With the development of traffic and the uprising speed of vehicle, the brake quality has been more concerned by peoples. High quality of brake is not only the guarantee to keep control of velocity, but also important to the safe driving. As the actuator of the brake system, the brake related directly to the automobile safety and the passengers'life. So how to design high quality and cheep brake based on modern design and manufacture technology is always the major problem concerned by all automobile factories and professional manufactories.While designing new brakes, the internal manufactories generally adjust the parameters by experience base on the known drawing of brake. As to the recurrence of calculation and the complexity of brake model, the process of the design can be only repeated in limited times manually, it results the parameters are rough. With the development of Parametric Modeling and Structural Optimization, the Parametric Optimization has been used widely in the engineering, and provided modern method to design automobile brake rapidly and delicately.However, as the brake model become complex gradually, the traditional algorithms make the Parametric Optimization more and more hardly. Fortunately, the parallelization trend of computer hardware and software brings us the parallel algorithm, such as Ant Colony Algorithm, and it provide new selection of Parametric Optimization Algorithm used in brake design. In this paper, the Ant Colony Algorithm has been researched, on the foundation of such work; the application of Continuous Ant Colony Algorithm in the brake parametric optimization has been done. The detail of the work is below:(1) As the Ant Colony Algorithm (ACA) is a new algorithm, the principle of Ant Colony Optimization has been analyzed, and both typical discrete ACA and continuous ACA have been implemented in MATLAB language. In order to calibrate parameters in algorithm, the Uniform Design Method has been used. Then a sick constrained function which is strong nonlinear has been tested both in CACA and Complex Algorithm which is typical traditional optimization algorithm, and the result shows that the CACA runs stably, the solution of the result is reliable. All of the work made the foundation of the applications in the parametric brake optimization.(2) Summed up the pressure distribution of the drum brake for various kindsof drum brakes'efficiency factor. Through the creation of simple model of vehicle brake course, some design requirements of Leading Trailing Shoe Brake of rear wheel has been asked. In consideration of self-locking criterion, lining wear criterion, size criterions and other performance criterions, the seven dimensions parametric optimization model of drum brake has been set up, which set the efficiency factor as the optimal object. The solution solved by CACA is better than the one which was solved by traditional algorithms and it all within the criterions.(3) On the foundation of summarizing the disc brake pressure distribution, the brake torque formulations of both floating caliper and complete disc were given. The modeling of the floating caliper brake of certain vehicle's front wheel takes in the following considerations: ten optimal criterions including brake torque, lining wear, uprising temperature and est., six design parameters including wheel cylinder diameter, drum diameter, brake line pressure and est., and there wear three optimal objects which are maximums brake torque, minimums disc thickness and minimums uprising temperature. In the process of resolving by CACA, normal form weighting method, exterior penalty function method, and linear mesh interpolation method were used to make CACA run effectively and efficiently. The result shows that most optimal objects have been improved.