BRIGE: A Fully Distributed Multi-Agent System Based Energy Management Solution for Microgrids

智能优点：智能技术是使现有电力系统更智能的具有成本效益的解决方案。为了提高微电网的可靠性和能效，降低控制系统的成本，采用全分布式控制方案是可取的。多智能体系统作为目前最流行的分布式控制解决方案之一，在电力系统运行控制中得到了广泛的应用。然而，现有的解决方案适用性有限，缺乏严格的稳定性分析。为了解决微电网的需求和现有解决方案存在的问题，S的长期研究目标之一是研究基于全分布式多智能体系统的控制解决方案。在本项目中，PI将设计三种基于稳定的全局信息发现算法的全分布式优化算法。所设计的算法可以根据模型的可用性、复杂性和响应时间等不同的能量管理任务进行选择。基于对代理的分析，将提出有效的同步技术。自主活动。将为具有多种不同类型分布式能源的微电网开发具有不同细节水平的模型。所开发的模型将用于评估候选解的动态性能，并克服优化算法的缺点。所提出的算法可以在几秒钟或更快的时间内实时解决简单问题或复杂问题。实时模拟和实验将被用来研究真实世界的表现。广泛的影响：该项目将通过引入新的理论和技术，在电力和能源研究中开辟新的领域。计划中的工作将向电力工程师展示先进的计算智能技术可以为复杂的工程问题带来新的解决方案。该项目的成功也将激励其他社会进行新的应用和挑战。该项目通过无缝集成电力系统调度和控制，有可能给电力系统运行实践带来革命性的变化。分布式优化算法可以应用于其他许多大规模的在线优化和控制问题。该项目将产生两个RA研究生职位，开设两个新的多学科课程，并通过为新墨西哥州州立大学的大量少数族裔学生提供绝佳的动手学习机会来扩大参与度。计划开展外联活动，以提高公众对智能电网的认识、理解和信心。根据该项目制作的微电网试验台、高质量的出版物和训练有素的学生将极大地提高新墨西哥州立大学进行相关研究和吸引政府和行业新资金的能力。

Intellectual Merit: Intelligent techniques are cost efficient solutions to making existing power systems smarter. To improve the reliability and energy efficiency of microgrids and lower the cost of the control system, fully distributed control solutions are preferable. As one of the most popular distributed control solutions, multi-agent systems have been widely applied for power system operation and control. However, existing solutions have limited applicability and lack rigorous stability analysis. To address the needs of microgrids and the problems with existing solutions, one of the PI?s long term research goals is to study fully distributed multi-agent system based control solutions. In this project, the PI will design three types of fully distributed optimization algorithms based on a stable global information discovery algorithm. The designed algorithms can be selected for different energy management tasks based on model availability, complexity, and response time, etc. Efficient synchronization techniques will be proposed based on analysis of agents? autonomous activities. Models with different levels of details will be developed for microgrids with multiple different types of distributed energy resources. The developed models will be used to evaluate the dynamic performances of candidate solutions and overcome the disadvantages of the optimization algorithms. The proposed algorithms can solve simple problems in real time or complex problems in several seconds or faster. Real-time simulation and experimentation will be used to investigate real world performance.Broader Impacts: This project will create new fields in power and energy research by introducing new theories and technologies. The planned work will demonstrate to power engineers that advanced computational intelligence techniques can bring about novel solutions for complex engineering problems. The success of this project will also inspire other societies with new applications and challenges. The project has potential to revolutionize the practice of power system operation by seamlessly integrating power system scheduling and control. The distributed optimization algorithms can be applied to many other large-scale online optimization and control problems. The project will generate two graduate RA positions, create two new multidisciplinary courses, and broaden participation by providing excellent hands-on learning opportunities for the large population of minority students at New Mexico State University. Outreach activities are planned to improve public awareness, understanding, and confidence of the Smart Grid. The microgrid testbed, high quality publications, and trained students produced according to this project will greatly enhance the capability of New Mexico State University in doing related research and attracting new funding from governments and industries.