EnICO – Energy efficient Industry Cluster Optimization

有限的资源，能源成本的上升和政府政策的不断变化，例如退出核能和化石燃料能源，对工业公司的经济分工施加了压力。此外，要求提高可持续性、减少环境污染和提高资源消耗效率的目的是改进对能源的处理。在此背景下，国家支持计划的重点是提高终端能源消耗的效率，特别是在消费者层面。相比之下，一个地方集群（例如工业园区）内的工业公司之间充满活力的相互依赖关系的潜力迄今尚未得到更深入的研究。因此，研究项目的主要目标包括集群内生产公司之间充满活力和实质性合作的建模和优化。根据资源效率的概念，产业集群的优化有两个主要目标：集群内部资源利用的最大化和外部资源需求的最小化。因此，计划开发一种优化工具（EnICO生成器），它将支持数字制图，以能源效率为目标的自我优化，并从长远来看，规划理想的生态工业园区。优化问题是离散-连续的，需要动态地将仿真与优化相结合。目的之一是确定可能对集群内能源需求产生积极影响的相互关系。第二个目标是达到能量最优的合作行为，从而提高能源效率。这是基于生成解决一组问题的解决方案，其中包括时间离散的、动态的和与顺序相关的有限资源分配。为此，相关关系的辨识和参数化、优化目标准则的建模和控制算法的定义是至关重要的。整体的项目目标导致了优化模型的高度复杂性。拟议项目的创新和有利因素之一将是映射的抽象层次，这对于模拟和优化单个企业之间充满活力的相互关系的效率至关重要。另一方面，通过实现多智能体系统（MAS）来保证智能集群优化。这个MAS系统地将“优化知识”结合到规则集，基于机器学习的方法。作为长期目标，应该能够生成特定的最优模型。为此，只有开发的规则集自动生成适当的相互关系到产业集群的基本模型中。

Limited resources, rise in energy costs and continuous change in government policies, e.g. exit from nuclear and fossil fuel energy, exert pressure on the economical division of industrial companies. In addition, the call for higher sustainability, the reduction in environmental pollution and efficiency in resource consumption aim for the improved handling of energy resources. In this context, the focus of national support programs pursue the increasing efficiency of end-use energy consumption, especially at the costumer level. In contrast, the potentials of energetic interdependencies between industrial companies within a local cluster (e.g. industrial parks) do not get on closer examination so far.Consequently, the primary objective of the research project comprise of both modelling and optimization of energetic and substantial cooperation between productive companies within a cluster. Following the resource efficiency concept, the optimization of industrial clusters has two main objectives – maximization of cluster internal resource utilization as well as minimization of external resource demand. Therefore, the development of an optimization tool (EnICO generator) is planned which will support the digital mapping, self-optimization with the goal of energy efficiency and in the long-term, the planning of ideal eco-industrial parks. The optimization problem is discrete-continuous and a dynamic combination of simulation and optimization is necessary. One aim is to identify the interrelations, which may have positive effects on energy resource demand within the cluster. The second aim is to reach an energetic-optimal cooperation behaviour, which increases the energy efficiency. This is based on the generation of solutions for solving the set of problems, which comprises time-discrete, dynamic and sequence-relevant allocation of restricted resources. For this, the identification as well as parametrization of relevant correlations, the modelling of optimization objective criteria and the definition of control algorithm are essential. The overall project objective leads to a high complexity of the optimization model. One of the innovative and advantageous factors of the proposed project will be the mapped level of abstraction, which is essential to model and optimize the efficiency of energetic interrelations between individual enterprises. On the other hand, intelligent cluster optimization is ensured by implementation of a multi agent system (MAS). This MAS systematically combines “optimization knowledge” to rule sets, based on the approach of Machine Learning. As long-term objective, it should be possible to generate a specific optimal model. For this, solely the developed set of rules automatically generate appropriate interrelations into a base model of an industrial cluster.