论文部分内容阅读
给出一种变阶式拟最小二乘法 ,该算法由主算法和切换算法组成 ,能在线估计模型的参数 ,当系统结构发生变化时 ,能自动改变模型阶数 ,并由切换算法为主算法提供初值 ,从而实现平滑过渡 .在此基础上提出了一种结构适应式自校正预报器 ,该预报器由参数估计器和最优预报器组成 ,能在线、实时、自动改变自身结构和参数 ,并能自动补偿由于模型不精确引起的误差 .用该预报器对某分馏塔温度进行 70步跟踪预报 ,平均相对预报误差为 2 .4 % ,预报精度提高了 0 .16 % .
This paper presents a quasi-least square method with variable order, which consists of the main algorithm and the switching algorithm, which can estimate the parameters of the model online. When the system structure changes, the model order can be changed automatically. Provide the initial value, so as to achieve a smooth transition.On this basis, a structural adaptive self-correcting predictor is proposed, which consists of parameter estimator and optimal predictor, which can change its structure and parameters automatically , And can automatically compensate for the error caused by the inaccurate model.With a 70-step tracking forecast of the temperature of a fractionator, the average relative prediction error is 2.4% and the forecast accuracy is improved by 0.16%.