In the field of sensor obstacle avoidance, the effect of measuring with a single sensor is not ideal. In practical applications, other types of sensors are often used for compensation, so as to achieve the best effect on the detection of the surrounding environment. Of course, this creates a problem of fusion processing of multi-sensor information, which increases the workload and difficulty of information processing. Then, in addition to this sensor obstacle avoidance method, there are many other methods to integrate and process a variety of sensor information, so that the full autonomous robot can achieve perfect obstacle avoidance, such as artificial potential field obstacle avoidance control method, fuzzy logic control obstacle avoidance control method, Artificial neural network obstacle avoidance control method, grid method obstacle avoidance control method and acoustic wave obstacle avoidance control method. Artificial potential field obstacle avoidance control method The artificial potential field obstacle avoidance control method is a relatively simple and novel approach. It is another kind of bionics. It follows the concept of potential and electric field force in physics, and establishes the virtual potential field in the robot workspace, according to the virtual potential field. The direction of force, to achieve local path planning. By constructing the artificial potential field of the gravitational field of the target pose and the repulsive field around the obstacle, the descending direction of the potential function is searched, and then the collision-free path is searched. It sounds very sinister, but there are already applications, and Khatib has been used in the navigation of mobile robots. But it has not been applied on a large scale. Because even for a simple environment, it is effective in static research, and does not consider the influence of obstacle speed and acceleration. Therefore, in dynamic obstacle avoidance control, artificial potential field obstacle avoidance control is not ideal. . Because in the complex multi-barrier environment, the unreasonable mathematical equation of the potential field is prone to local extremum points, causing the robot to stop moving when it does not reach the target, or to generate oscillations, swings, and the like. In addition, the traditional artificial potential field method focuses on obtaining a feasible path that can avoid obstacles, and has not yet studied what optimal path. Fuzzy logic control obstacle avoidance method The fuzzy logic control obstacle avoidance method appeared not too late. In 1965, a professor in the United States proposed the concept of fuzzy logic. In 1974, University of London, a professor of fuzzy control statements fuzzy controller controls operation of the boiler and gas turbine to be successful, start the fuzzy math used in automatic control fields, including the field of robotics. Since there is no need to create an analyzable environment model, the current fuzzy logic method has a lot of research work in solving the problem of robots avoiding obstacles. Another unique advantage also makes it possible to adjust the rules with expert knowledge, because each rule of the rule base has a clear physical meaning. In the fuzzy logic control obstacle avoidance method, the fuzzy control rule is the core of fuzzy control. One of the new trends in current research work is its increasing nature, especially in the automatic generation of fuzzy control rules, that is, along with automatic fuzzy data acquisition, the algorithm provides online fuzzy rule learning ability, data acquisition and rule generation are automatically executed. Artificial neural network obstacle avoidance control method Artificial neural network is a network system with parallel computing ability, which is composed of many units (also called neurons) connected according to a certain topological structure. It has strong nonlinear fitting ability and multiple input and multiple output simultaneously. The ability to handle. Used in robots, it is to simulate the human brain neural network to process information, and to obtain information processing capabilities like human brain from another research perspective. For intelligent robots , the use of artificial neural networks for information fusion has one of the greatest advantages, namely, large-scale parallel processing and distributed information storage, with good self-adaptation, self-organization, and strong learning functions and association functions. And fault-tolerant function, close to the information processing mode of the human brain. China Portable 240V Ev Charger,Portable Electric Car Charger supplier & manufacturer, offer low price, high quality Portable Electric Vehicle Charger,Portable Level 2 Charger, etc. Portable 240V Ev Charger,Portable Electric Car Charger,Portable Electric Vehicle Charger,Portable Level 2 Charger Shenzhen Hongjiali New Energy Co., Ltd. , https://www.hjlcharger.com
