|
所在平台: Udemy |
课程主页: https://www.udemy.com/course/modeling-simulation-and-control-using-python/
课程评论：没有评论
课程名称：使用Python进行建模、仿真与控制 课程概述： 本课程针对系统建模进行讲解，涉及以不同形式创建系统的抽象模型，如微分代数方程、框图表示、传递函数、状态空间表示及信号流图等。系统建模的目的是为了使模型的动态特性和信号流动能够被理解，从而进行控制操作，并预测系统在参数变化下的行为。课程重点关注离散时间域中的模型。 离散时间域模型： 该模型的操作不是连续的，而是针对每个时间步进行离散计算。通过收集系统数据，并应用数学/物理定律，将其转化为微分方程来创建系统模型。离散模型包含输出随时间变化的幅度。为了改变系统的输出，可以调整系统的输入。控制理论是一门研究选择适当输入以产生期望结果的策略的数学分支。 仿真： 在软件（Python）中对模型进行仿真时，需要设置时间步长、结束时间、积分器/求解器等环境参数。系统输入以系统化的方式在选定的时间范围内给出，并使用适当的步长。这种仿真方式为选定输入在时间范围内生成输出。 控制： 应用一种策略测量当前输入的输出，并将其与参考/期望输出进行比较，利用此差异（误差）项在下一步（未来）更改输入，直到结束时间。如果下一步的误差在基于当前输入的情况下能够减少，则系统趋近稳定解；否则，控制策略不满足稳定性标准，需要进行修改以减少未来的误差。 学习内容： 1. 系统建模 - 为三个例子形成微分代数方程 2. 系统仿真 - 使用Python构建模型和仿真设置 - 针对三个实例 3. 系统控制 - 应用控制策略，通过根据误差项选择适当输入来将输出驱动朝向目标。 本课程适合希望深入了解系统建模、仿真与控制的学习者。
System modeling deals with the creation of abstract models of system in different forms such as Differential algebraic equations, Block diagram representation, Transfer functions, State space representation, signal flow graph, etc. It is used to present the model so that the dynamics of the model and the flow of signals within is understandable and further control operations can be performed and behavior of the system can be predicted for change in parameters.In this course we deal with discrete models in time domain.Discrete models in time domain:The model whose operation is not continuous, but discrete for each time steps. Calculation is performed for each time step interval. System model is created by collecting the system data and applying mathematical / physical laws and transforming them into differential equations. Discrete model contains the amplitude of the output with respect to time. To vary the output of the system, the input to the system can be varied. The branch of mathematics that deals with strategy to select the appropriate input to generate the expected outcome is control theory.The model to be developed should posses the output as function of input. The function includes all the possible/realistic dynamics. The same model can also be represented in frequency domain which involves amplitude of the output with respect to time.Simulation:Simulating the model involves creating it in the software (Python) and setting up the environment parameters like time step, end time, integrator/solver, etc. The input to the model is given in a systematic way for the selected time range with appropriate step size. This type of simulation creates the output for the given input across the time range.Control:Applying a strategy which measures output for current input and compares it with reference/desired output and uses this difference(error) term to change the input to the system in the next step (future) this process continues until the end time. If the error in next step is going to reduce based on the current input, then the system is approaching stable solution. Otherwise, the control strategy does not meet stability criteria and is not the proper one. Control strategy has to be modified to reduce future errors.In this course, the following can be learnt:1. System modelling - Form differential algebraic equations for system - for three examples2. System simulation - building the model and simulation set up in python - for three examples3. System Control - applying control strategy which involves drives output towards objective by selecting appropriate input based on error term.
