Research On Chinese Instrction Task Parser Issues For Home Service Robot

In the field of home service robot research,it is one of important research topic in improving the intelligence and independence of robot that how to transform Chinese instruction in to a sequence of actions which robot can understand.To address the problem of home service robot which can autonomously execute service task contained in the Chinese instruction,and robot which can autonomously deal with Chinese instruction without needing artificial interference,this paper presents a task parser method based on improved Answer Set Programming for Chinese service instructions,and utilize methods like nature language processing,task planning,and cost planning to deeply studies the above-mentioned issues.First of all,the Chinese word segmentation and task planning methods used in this paper are elaborated in detail.The Institute of Computing Technology,Chinese Lexical Analysis System is briefly introduced,and the method of chunk labeling,answer set programming related theoretical knowledge,syntax semantics and corresponding solver are analyzed.Secondly,the task planning module maps the Chinese instructions to the best action sequences that the robot can execute based on the environment description,the initial state,the goal state,and atomic action related content.Based on this,joining the cost in planning task,calculate the cost of the action sequence and choose the answer set with the lowest cost as the optimal action sequence.In addition,ask actions and sense actions are added to improve scene information through human-robot interaction,which can quickly and efficiently obtain the required correct information,thereby improving the efficiency of service instruction task resolution.Thirdly,task planning modules do research about robot task planning.This module will mapping Chinese instruction to optimal active sequence of robot could manipulate by environment description,the original state and the object state and the relative contents of atom active.On this basis,the paper adds cost planning,calculate the cost of active sequence,and the minimum-cost Answer Set is chosen as the optimal active sequence.And join ask active and sense active,to pass Human-Robot Interaction to improvement environment information.Get the right information faster and more efficiently,and then improve the efficiency of task planning.Finally,simulates the experiment on the service robot platform,and gives the experimental results of each link in Chinese service instruction processing.The mapping process from the Chinese service instructions to the executable action sequence was shown with examples.Finally,the feasibility and correctness of this strategy is verified with the experimental results of task planning and cost planning of Chinese instruction.