In this paper, by integrating neural network approximators, a data-based composite control technique is developed with critic learning implementation and wastewater treatment verification. The iterative adaptive critic framework is established involving dual heuristic dynamic programming (DHP), so as to obtain an intelligent optimal controller. Besides, a steady control input is computed with the help of the neural identifier. Then, by combining the DHP controller and the steady control input, an effective composite control strategy is derived and applied to the proposed wastewater treatment platform. Through conducting experiments, it is observed that the dissolved oxygen concentration and the nitrate level can be maintained at setting points successfully, which results in an intelligent wastewater treatment system.

中文翻译：

---

具有批判智能的污水处理平台基于数据的复合控制设计

在本文中，通过集成神经网络逼近器，开发了一种基于数据的复合控制技术，并结合了批评学习实施和废水处理验证。建立迭代自适应批评框架，涉及对偶启发式动态规划（DHP），以获得智能最优控制器。此外，在神经识别器的帮助下计算稳定的控制输入。然后，通过结合 DHP 控制器和稳定控制输入，推导出有效的复合控制策略并将其应用于所提出的废水处理平台。通过进行实验，观察到溶解氧浓度和硝酸盐水平可以成功地保持在设定点，从而形成智能废水处理系统。