毕业设计&课设-基于MATLAB的过程控制仿真.zip

# Process Control Simulations (PCS) for MATLAB **PCS** is an open-source object-oriented library developed by Departamento de Informática y Automática at UNED, aiming to provide an easy-to-develop framework for process control simulations. ## Table of Contents - [Introduction](#introduction) - [Getting Started](#getting-started) - [Using the Library](#using-the-library) - [API Reference](#api-reference) ## Introduction This manual describes the use of Matlab-based simulation library PCS, which facilitates development of process control loops in networked and/or event-based environments. Several tutorial examples are provided to show how to prepare reusable controllers and processes and perform continuous, periodic or event-triggered simulations. This manual also includes a detailed API reference of the library with every package and its contents. For questions and bug reports, please direct them to the [issues page of the library](https://github.com/crcuned/pcsmatlab/issues). ## Getting Started Download the zip archive from [PCS for MATLAB homepage](https://github.com/crcuned/pcsmatlab). Then 1. Extract all files to some suitable directory. 2. Start MATLAB R2014a or higher and navigate to previous directory. 3. Run *install.m* to add the necessary directories to the MATLAB path. ## Using the Library Tutorials are given in ascending difficulty order. ### Development of a Quadruple-Tank Process The following class represents the quadruple-tank process described [here](http://ieeexplore.ieee.org/document/845876/). ~~~matlab classdef QuadrupleTank < PCS.Process.Process properties a % Cross-sections of the outlets holes A % Cross-sections of the tanks g % Acceleration of gravity gamma % Positions of the valves k % Voltage to volumetric flow rate gains end methods function self = QuadrupleTank(a, A, g, gamma, k) if ndims(a) ~= 2 || length(a) ~= 4 error('Cross-sections of the outlets holes must be a vector of length 4.'); end if ndims(A) ~= 2 || length(A) ~= 4 error('Cross-sections of the tanks must be a vector of length 4.'); end if ~isscalar(g) error('Acceleration of gravity must be a scalar.'); end if ndims(gamma) ~= 2 || length(gamma) ~= 2 error('Positions of the valves must be a vector of length 2.'); end if ndims(k) ~= 2 || length(k) ~= 2 error('Voltage to volumetric flow rate gains must be a vector of length 2.'); end self.a = a; self.A = A; self.g = g; self.gamma = gamma; self.k = k; self.n_inputs = 2; self.n_outputs = 2; self.n_states = 4; end function dxdt = derivatives(self, t, x, u, ~, ~) dxdt = zeros(4, 1); xt = x(t); dxdt(1) = -self.a(1)/self.A(1)*sqrt(2*self.g*xt(1)) + self.a(3)/self.A(1)*sqrt(2*self.g*xt(3)) + self.gamma(1)*self.k(1)/self.A(1)*u(t,1); dxdt(2) = -self.a(2)/self.A(2)*sqrt(2*self.g*xt(2)) + self.a(4)/self.A(2)*sqrt(2*self.g*xt(4)) + self.gamma(2)*self.k(2)/self.A(2)*u(t,2); dxdt(3) = -self.a(3)/self.A(3)*sqrt(2*self.g*xt(3)) + (1-self.gamma(2))*self.k(2)/self.A(3)*u(t,2); dxdt(4) = -self.a(4)/self.A(4)*sqrt(2*self.g*xt(4)) + (1-self.gamma(1))*self.k(1)/self.A(4)*u(t,1); end function y = outputs(self, t, x, u, ~, ~) y = x(t, [1 2]); end end end ~~~ ### Development of a PI Controller The following class represents a general PI controller. ~~~matlab classdef PI < PCS.Control.Controller properties K % Proportional gains Ti % Integral times end methods function self = PI(K, Ti) if ndims(K) > 2 error('Proportional gains must be a vector.'); end if ndims(Ti) > 2 error('Integral times must be a vector.'); end if length(K) ~= length(Ti) error('The lengths of the proportional gains and integral times must match.'); end self.K = K; self.Ti = Ti; self.n_inputs = length(K); self.n_outputs = length(K); self.n_states = length(K); end function dxcdt = derivatives(self, t, ~, ~, y, ~, r) dxcdt = self.Ti.^(-1).*(r(t) - y(t)); end function u = outputs(self, t, xc, ~, y, ~, r) u = self.K.*xc(t) + self.K.*(r(t) - y(t)); end end end ~~~ ### Executing a Simulation The following script is used to perform the [PI control](#development-of-a-pi-controller) of the [quadruple-tank process](#development-of-a-quadruple-tank-process). ~~~matlab %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % PI control of a quadruple-tank process % %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % Quadruple-tank process a = [0.071 0.057 0.071 0.057]; A = [28 32 28 32]; g = 981; gamma = [0.7 0.6]; k = [3.33 3.35]; process = QuadrupleTank(a, A, g, gamma, k); % PI controller K = [0.3816; 0.5058]; Ti = [62.9557; 91.3960]; controller = PI(K, Ti); % Create simulation simulation = PCS.Simulation(controller, process); % Define initial states and time interval simulation.xc0 = [31.4347; 33.4446]; simulation.x0 = [12.4; 12.7; 1.5919; 1.4551]; simulation.t0 = 0; simulation.tend = 500; % Define set-point conditions simulation.set_preloaded_reference(1, 15); simulation.set_preloaded_reference(2, 12.7); % Execute simulation data = simulation.run(); % Plot results plot(data.t, data.x); ~~~ ## API Reference The API reference is organized by packages. | Package | Description | |:----------------------------- |:------------------------------------------------------------------------------------------------------------ | | [PCS](#pcs) | Provides the core classes necessary to perform a process control simulation and the base system definition. | | [PCS.Control](#pcscontrol) | Provides the base controller definition and some of the well-known control laws. | | [PCS.Hardware](#pcshardware) | Provides classes for emulating physical interaction with real processes. | | [PCS.Network](#pcsnetwork) | Provides classes for establishing a networked communication between control loop components. | | [PCS.Process](#pcsprocess) | Provides the base process definition and some of the well-known industrial processes. | | [PCS.Utils](#pcsutils) | Provides static classes with useful reusable functions to simplify the simulation of the closed-loop system. | ### PCS Provides the core classes necessary to perform a process control simulation and the base system definition. #### Class Summary | Class | Description | |:---------- |:----------- | | [Simulation](#class-simulation-pcs) | The *Simulation* class contains the core functionality of the library that allows to perform process control simulations. | | [System](#class-system-pcs) | An abstract class which provides a common definition for processes and controllers. | ##### Class Simulation ([PCS](#pcs)) - [handle](https://es.mathworks.com/help/matlab/ref/handle-class.html) The *Simulation* class contains the core functionality of the library that allows to perform process control simulations. ###### Properties Summary | Modifier | Property | Description | |:-------- |:--------- |:----------- | | None | actuators | Vector containing an actuator for each input of the process. | | None | actuator\_links | Vector containing the links of the network controller-process (C-P). | | None | d\_sensors | Vector containing a sensor for each disturbance affecting the process. | | None | event\_tol | Zero-crossing tolerance for event location. | | None | on\_zeno\_enter | Function handle to be called before a Zeno phenomenom occurs. | | None | on\_zeno\_exit | Function handle to be called after Zeno phenomenom is solved. | | None | sensor\_links | Vector containing the links of the network process-controller (P-C). | | None | solver | A function handle to the ODE solver. | | None | t0 | The ini