Research And Implementation Of Two-wheel Balance Intelligent Vehicle Following Design

The two-wheel self-balancing follow-up vehicle can be regarded as an intelligent robot system with two-wheel parallel layout structure.It breaks through the concept of traditional car.It is also set attitude detection and acquisition,feature target recognition and positioning,behavior state control and execution.A complete system with innovative features such as flexibility,self-balancing,and specific follow-up.As a pioneer of technological innovation in the new era,the two-wheeled manned balance car has begun to expand from the field of manned self-balancing electric vehicles.With the development of service-oriented intelligence and meeting people's higher expectations for the performance of this product,the application of development and expansion has gradually increased,and gradually appeared in the consumer market.Under this premise,based on the research on the key technology of the mature manned balance electric vehicle in our laboratory,based on the self-balancing manned electric vehicle,the research and innovation has developed a monocular vision following the special target two-dimensional code's system.Therefore,this paper mainly does the following learning and research work:1.A relatively complete two-wheel self-balancing robot mathematical model is established,and the feasibility of two-wheel self-balancing and following dynamic balance is analyzed.At the same time,the simulation results of the following control theory are obtained by adding the set control amount.Through this result,the algorithm design work of the software control part and the parameter setting inside the algorithm are guided.2.The following techniques and precisions in many literatures are analyzed.A two-wheel self-balancing robot following scheme with monocular vision following target code is selected,that is,a monocular vision module with STM32F7 high-performance controller is selected and built.A relatively complete and suitable two-wheel self-balancing hardware device is established,and the software control scheme of the three-loop servo PID algorithm is determined.3.A set of preprocessing algorithms based on the combination of judgment method,threshold method and sliding window method to capture the visual target following the effectivetarget is studied.At the same time,the situation after losing the target is adopted by using the fifth-order least squares curve.The algorithm analyzes and predicts the history of 36 sets of coordinate data,and establishes a set of prediction and re-pickup strategies after the target is lost in the following process.4.For the transplant following function on the manned balance vehicle,the hardware system of the manned balance vehicle is optimized,and the motor control problem in the transplant process is solved through research,and the following effect of the manned balance car is basically realized.Then,the two system platforms of the two-wheel balance following vehicle and the man following the balance vehicle were tested respectively,and the effects of self-balancing stability,following anti-interference,following stability and regaining strategy after loss were tested.5.Finally,through the analysis of the test results and the summary of the thesis system,the shortcomings of the work and the direction of the expected research and improvement are put forward,and the writing of the whole paper is completed.