This paper introduces An Intelligent Agent Model for a Given Task in a Specified Environment. The methodology adopted in this work is based on mixing computational methods and functions to build an intelligent agent model. This paper focuses on building an intelligent agent model as a knowledge-based system that interacts with a dynamic environment for performing tasks. The class structure used to represent the environment in the knowledge base relies on three types of knowledge representation forms: production rule, semantic net, and frames. Each object in the environment is an instance of the class environment. Algorithms and functions are used to get knowledge from the state space of an environment to construct a task. The intelligent agent model can understand the environment from any position and can detect many subtasks, arrange them in a queue for execution, and can make decisions at a high scale of thinking. This model is proposed to maintain that an agent which is characterized by sufficiently low computational costs can interact with the environment in real-time but is powerful enough to reach the assigned goals in complex environments and within an acceptable time period. The intelligent agent model can calculate persistent changes in an external dynamic environment and any unexpected change, for example detecting the being of any problem in the environment and avoiding it. The intelligent agent can also learn and take reasonable decisions in the dynamic environment and automatically select an action based on task features. Thus, the intelligent agent can resolve several different kinds of difficulties.
展开▼
