动态昂贵优化算法及其在供水管网突发性污染源实时定位中的应用

In terms of solving the problems of drinking water contamination caused by accidental contaminant event in water supply network, we used the techniques just like sensor network which could identify the source location to isolate the contaminated area and minimized its hazards. It was significant to the drinking water security emergency handing. How to use the detecting information of the sensor to identify the contaminant source is an important multi-disciplinary problem. Due to the problems with sudden, time-varying and timeliness characteristics, so the sudden contaminations source of real-time identification in water supply network is a typical real-time, dynamic and expensive optimization problem. We plan to focus on the fundamental theory, optimization model and algorithm : (1) Based on hydraulic, water quality model and theory of network dynamics to quantify the problem of sudden contaminations source of real-time identification in water supply network. (2) Using the Gaussian model to simulate the dynamic change of the water demand of users, establish a dynamic optimization model to study the uncertainty of solution. (3) In order to solve the computational expensive problem in sudden contaminations source of real-time identification , a dynamic expensive optimization algorithm based on Gaussian Agent model is proposed and implemented by modifying the adaptability of expensive optimization model. Through the research of our project, can be explored to solve the basic theory and method of dynamic expensive optimization problems, is of great scientific significance and application value.

针对供水管网中突发污染事件引起的饮用水污染问题，利用传感器网络等技术手段快速准确地定位污染源位置及其影响范围，为饮用水安全保障提供有力的技术支持，具有重要的现实意义。如何利用传感器数据，迅速判断污染源的可能位置，是一个多学科交叉问题。由于该问题具有突发性、时变性和时效性等特点，属于典型的实时、动态和昂贵优化问题。本项目拟针对此问题的基础理论、优化模型和算法进行研究：(1) 基于水力水质传输模型和管网动力学理论，从系统层面量化供水管网突发性污染源实时定位问题；(2)利用高斯模型模拟用户水需求动态变化的规律，建立污染源实时定位的动态优化模型，研究解的不确定性问题；(3)对昂贵优化模型进行适应性修改，提出并实现基于高斯代理模型的动态昂贵优化算法，解决突发性污染源实时定位中的计算昂贵性问题。通过本课题的研究，可以探索求解动态昂贵优化问题的基础理论及方法，具有重要的科学意义与应用价值。

针对供水管网中突发污染事件引起的饮用水污染问题，利用传感器网络等技术手段快速准确地定位污染源位置及其影响范围，为饮用水安全保障提供有力的技术支持，具有重要的现实意义。如何利用传感器数据，迅速判断污染源的准确位置，是一个多学科交叉问题。由于该问题具有突发性、时变性和时效性等特点，属于典型的实时、动态和昂贵优化问题。本项目针对此问题的基础理论、优化模型和算法进行研究：(1)基于水力水质传输模型和管网动力学理论，从系统层面量化供水管网突发性污染源实时定位问题并提出了相应的优化模型；(2)使用高斯模型、泊松分布模型和自回归模型模拟用户水需求动态变化的规律，建立了污染源实时定位的动态优化模型，并提出了问题求解的相应算法；(3)对昂贵优化模型进行适应性修改，提出并实现了基于高斯代理模型的动态昂贵优化算法，解决了突发性污染源实时定位中的计算昂贵性问题。通过本课题的研究，可以探索求解动态昂贵优化问题的基础理论及方法，具有重要的科学意义与应用价值。

内容获取失败，请点击重试