Design And Implementation Of Configuration Element Of Single Neuron PID

The single neuron PID control algorithm, using the neuron network theory to improve the conventional PID algorithm, is a kind of typical advanced control strategy, which has better control performances, including less adjustable parameter, easier to adjust, better control performance to large delay object and easier to be realized in industry. By far, this algorithm has been successfully applied in various industrial control systems. However, the implementation methods are different and the controllers have strong specificity, which lead to the poor universality and portability of algorithm, and will be limited to all kinds of hardware conditions and control platforms when used in real application. By using single neuron PID algorithm used in practical control system as entry point, and adopting IAP (Industry Automation Platform) as environmental platform, which is a new generation distributed control platform developed by Fuzhou Histron Research Institute, this paper adopts graphical configuration technology to research the thinking and methods of using graphical configuration components to realize single neuron PID algorithm. Besides, through control configuration packaging techniq ue, this paper completes the design of single neuron PID configuration components.The primary research works are as follows:Firstly, this paper introduces the control configuration technology and analyzes the principle and character of IAP platform modeling, so as to provide theoretical basis for realizing the control algorithm of configuration components. Secondly, this paper analyzes structure of single neuron PID algorithm, designs single neuron PID algorithm in IAP platform, and finally realizes single neuron PID algorithm by configuration components. Thirdly, combining the characters of control configuration programming on IAP platform, this paper uses configuration components to write online analytical algorithm of control configuration, which can monitor and calculate the steady-state and transient performance of the step response of the control system in real time while the system is running. Fourthly, the paper sets up a three order model as a controlled object in order to build up the single neuron PID and convention PID control system, and the two control systems will be downloaded to simulation control station at the same time. Using the functions of online configuration and forcing value provided by IAP platform, the controller can be debugged online and the online configuration analyzer can quantitatively compare the control performances of the two controllers so as to prove the superiority of single neuron PID control algorithm over conventional PID. At last, through IAP platform control configuration packaging technique, a design method which encapsulates with single neuron PID control configuration logic is proposed, and the relationship between the structure of single neuron PID algorithm and packaged components on IAP platform is discussed. Through component packaging technique, a set of new control components are created for single neuron PID algorithm, which realizes the single neuron PID algorithm. The structure is close to the topological graph algorithm, which has good readability and is liable to realize. Compared with the result operated in industrial control station, it’s found that the algorithm performance is just the same with basic configuration method.The development process for single neuron PID can also be applied to other advanced control algorithms, and this paper can be used as a universal way and thought for industrial application of advanced control technology. It has pracitcal significance in academic research and engineering applications.