Autonomous Decentralized Manufacturing System Research Articles

An Agent Behavior Technique in an Autonomous Decentralized Manufacturing System

In this paper, we propose a technique for the communication between agents in the dynamic generation of a production plan in an autonomous production system using product agents, parts agents and assembly machine agents. A parts agent leads the promotion of processes in this type of assembly system. First, product agents create the assembly process model that contains the assembly sequence to complete the product. Next, product agents generate parts agents that have the assembly process model. Parts agents will be assembled by machines, with some communications between other agents. The proposed system is defined as an event driven architecture. The increase in the costs of communication between agents is polynomial in the number of the agents. Moreover, this system adjusts the production schedule dynamically using only local negotiation when conditions are changed. Finally, we show experimental results with simply examples.

3304 自律型生産システムのためのエージェントの制御方法(OS3-3 先進生産システム)

3304 自律型生産システムのためのエージェントの制御方法(OS3-3 先進生産システム)

1104 製品エージェントに基づいた生産システムの構成とスケジューリング法(OS1 生産システムの設計・管理・評価)

In this research, a product agent is proposed for the configuration of an autonomous decentralized manufacturing system which flexibly responds to high-mix very-low volume production. A product agent has its own product model. A product agent can select the machine to manufacture the product by referring to machine data and negotiation with other product agents. When the material for the product enters the production line, the corresponding product agent is dynamically generated. The agent confers with other agents on the production machine to be used for each job. The method of achieving the negotiation was examined by generating a production schedule. By this trial implementation, the advantage of the proposed agent-negotiation method was shown.

Intelligent AGV driving toward an autonomous decentralized manufacturing system

This paper proposes two methods that give intelligence to automatically guided vehicles (AGVs). In order to drive AGVs autonomously, two types of problems need to be overcome. They are the AGV navigation problem and collision avoidance problem. The first problem has been well known since 1980s. A new method based on the feature scene recognition and acquisition is proposed. The sparse distributed memory neural network (SDM) is employed for the scene recognition and acquisition. The navigation route for the AGV is learnt by use of Q-learning depending on the recognized and acquired scenes. The second problem is described as mutual understanding of behaviors between AGVs. The method of mutual understanding is proposed by the use of Q-learning. Those two methods are combined together for driving plural AGVs autonomously to deliver raw materials between machine tools in a factory. They are incorporated into the AGVs as the machine intelligence. In experimental simulations, it is verified that the first proposed method can guide the AGV to the suitable navigation and that the second method can acquire knowledge of mutual understanding of the AGVs’ behaviors.

AGV Autonornous Driving Based on Scene Recognition by Use of the Simplified SDM.

An intelligent material-handling system is an important part of an autonomous decentralized manufacturing system(ADMS). Especially, an automatically guided vehicle(AGV)is the core of the material-handling system. This paper reports on a method to drive the AGV autonomously in the ADMS. The new method proposed here combines the sparse distributed memory neural network(SDM)with Q-learnig(Q-L). The SDM explores and acquires feature scenes required for AGV driving. The Q-L finds a direction to take at a particular scene acquired by the SDM. Numerical simulation verifies that the SDM can extract the feature scenes necessary for driving the AGV, and that the Q-L can instruct the right direction to the AGV at each extracted scene so that the vehicle may move towards the target location.

Practical study on wireless transmission of power and information for autonomous decentralized manufacturing system

In previous papers, the authors have already introduced the idea of an autonomous decentralized manufacturing system (ADMS), in which machines receive electrical energy through wireless power transmission and also are decentrally controlled through wireless transmission of the control and feedback data. With this idea, the degree of freedom in motion control is enlarged, and the system has the capability of adaptive reconfiguration to product variation as a result. In this paper, the practical problems for realization of wireless transmission of power and information (WTPI) needed for the proposed ADMS are clarified, and the practical solutions to these problems are presented. Several industrial applications of the WTPI are also introduced.

Studies of Autonomous Decentralized Manufacturing System. 1st Report, Optimization of Single-Period Production Planning Problem by Noncooperative Game.

In this paper, a method for the decentralization of multi-stage manufacturing systems is proposed under the noncooperative game. Necessary optimality conditions and Nash equilibrium solution as optimal production plan are derived for single period production planning problem of autonomous decentralized manufactuing systems that consists of multistages. Numerical examples are solved to examine the validity of the proposed approach.