کنترل تطبیقی ​​زنجیره تامین: بلوک های ساختمان و ابزار بر اساس یک عامل چارچوب شبیه سازی

An adaptive control concept for supply chains is presented. Its background is a complex supply chain scenario originating from the machine building domain with a focus primarily on the development of an agentbased method concerning the adaptive coordination of customer orders along the supply chain. It aims at flexibly handling disturbances in relation to the re-allocation of alternative suppliers to ensure a timely and accurate fulfilment of customer orders. The researched and described building blocks and tools originate from artificial intelligence, decision theory and operations management, which have been implemented in an agentbased simulation framework.

Companies are forced by competition to develop new optimization potentials. After nearly all internal possibilities of companies to improve have been exhausted, the off-site concepts to generate competitive advantages seem to be very promising. In view of this situation, the supply chain management concept appears to be an appropriate solution. Therefore, the main goal is to organise the overall process of the supply chain, especially the top-level order processes, to achieve an optimum in cost and time. This can be achieved, for example, by improving key figures,such as increasing capacity utilization, decreasing inventories, decreasing lead times and so on. The supply chain represents a very dynamic and complex multi-level system with multiple input and output relations [1]. In this kind of tight networked system, stability frequently rises to a certain degree, before it declines again due to the greater complexity. In this context, supply chain management can be regarded as a management or control concept for complex, dynamic systems, which imperatively needs to be reflected in the development of modern planning and control concepts [1, 2]. Today’s conventional control of a supply chain mainly pursuits sequential, top-down planning approaches supported by different MRP (Material Requirements Planning) or ERP (Enterprise Resource Planning) concepts and information technologies. They coordinate the supply flows between the different companies, which very often cause time lags [3]. Sudden disturbances in the supply chain basically ripple all the way through and can easily make the complex and inherently distributed planning processes invalid. Expensive re-planning sessions about the quantity delivered, the delivery times and search for new suppliers are the most likely consequences. Taking these aspects into account, a concise supply chain management concept for tomorrows near future must clearly address the following, very challenging key issues: • Acceleration and especially direct transmission of information flows related to customer orders starting at the point of sale • A deeper understanding of the relation between the product structure and the structure (customer order oriented) configuration of the supply chain respectively • More efficient methods to support a coordinated, decentralised and pull-oriented scheduling • Enriched and extended decision [4] and planning methods for an efficient and effective handling of disturbances along the supply chain • Intelligent support concerning the re allocation of alternative suppliers [5] on all levels of the supply chain in order to be able to fulfil the customer orders at any time The following sections present some of the most promising basic building blocks for an adaptive supply chain management against the background of a conducted research project and on the basis of an underlying industrial scenario from the machine building industry.

Building blocks for the design of an adaptive control architecture for supply chain management was the topic being presented in this paper. It was conveyed in terms of deployed methods, theories and tools, against the background of a previous research project. On the basis of an underlying industrial scenario, a multi-agent based supply chain control module was developed. It was connected to a simulation framework to primarily display and verify the dependencies between the communication and planning of the software-agents in relation to the material flows in the supply chain model. It concerned a customer oriented processing of orders and an adaptive re-allocation of suppliers,due to unforeseen disturbances. The presented agent-based order coordination process can not be compared to any currently available commercial tools, e.g. to scheduling suites. A direct bench marking, therefore, is presently not possible.