Intelligent Greenhouse Control System Based On Multivariable Control

Greenhouse environment system is a nonlinear, time-varying, time-delay and the complexlarge system, it is difficult to establish mathematical model of the system and there are a lot ofdisturbances in the system, and the process object contains temperature and humidity is stronglycoupling and large time-delay. Simple single-input single-output control strategy and thetraditional PID control algorithm in intelligent greenhouse control system of large lag processoften have oscillation, overshoot phenomena and poor robustness. In this paper, according to thecharacteristics of the greenhouse environment control, a predictive PI algorithm and dynamicdecoupling control designing scheme based on multivariable control is proposed in this paper,when a better change is used to the traditional decoupling control system, and add predictive PIcontrol algorithm to theoretical research strategy. After online simulation, the results showdecoupling control strategy based on predictive PI has better stability, robustness,anti-disturbances ability, and practical application value. Its main content is divided into thefollowing three parts:1. Aiming at the strong coupling process, a kind of advanced dynamic decoupling controlalgorithm is studied and proposed. The basic idea is dynamically decoupling firstly, thensimplifying decoupled object as independent first-order lag objects, at last using single variablealgorithm designing the first-order lag object’s controller. This algorithm is compared withtraditional decoupling control algorithm, its main characteristic is the dynamic decouplingalgorithm is easy to implement and can be the organic combination of advanced control algorithm.2. For the big lag process greenhouses, according to the desired closed-loop transfer function,studying and putting forward a kind of advanced predictive PI controller, this controller has twoparts, one has the form of PI controller, another has the forecast function. This kind of controllerhas a simple structure, less adjustable parameters, the adjustment of the parameters is simple andintuitive, the simulation shows that it has good tracking performance and good anti-interferenceperformance, and robust stability performance is good, suitable for promotion and application inthe industrial production. 3. Studying the advanced control algorithm in the application of the multivariate intelligentgreenhouses, the advanced control algorithms of this paper including dynamic decoupling controlalgorithm and the predictive PI control algorithm. First according to the characteristics of thegreenhouses temperature and humidity object process, the temperature and humidity processobject model is established, respectively. The process model are included in the first-ordertime-delay object, firstly using dynamic decoupling controller for the phenomenon of strongcoupling. Then we design predictive PI controller of using the predictive PI control algorithm forthe process object model of temperature and humidity respectively. In addition, also theintelligence greenhouse control scheme is analyzed in this article, the actual control effect showsthat the advanced control algorithm can not only improve control accuracy, high stability and havestrong robustness,.4. In order to study the engineering algorithm, the real-time monitoring software designbased on OPTO22PAC Project is designed in this paper, by using MATLAB SIMULINK moduleto simulate the actual process object, studying the performance of the monitoring and controlsoftware through simulation. And we also develop a real-time monitoring of human-machineinterface(HMI), users can directly on the control interface change the various parameters of thecontroller, it can clear seeing that the change of the waveform, the human-machine interface(HMI)is simple, convenient and practical.