Method and system for evaluating response of entity robot based on deep learning

Abstract

The invention provides a method for evaluating the response of an entity robot based on deep learning. The method comprises the following steps: S1, monitoring the condition of each entity robot, andpushing corresponding robot equipment information to a user; establishing an evaluation task, selecting a to-be-evaluated robot, and setting response time, frequency and duration of an execution linkof the to-be-evaluated robot; S2, monitoring whether a to-be-evaluated task exists or not, and if so, initiating an evaluation execution request according to evaluation information issued by the evaluation task; S3, storing the evaluation data of each robot, and recording the data by taking the entity robot and each test as unique identifiers; S4, performing data analysis, evaluation result statistics and evaluation coverage rate on an evaluation result, and displaying the evaluation result in a graphical mode; and S5, performing deep learning for each test mode and the output test data, and optimizing the test mode and the test report; evaluation of the response speed of each capability of the entity robot is completed, and the credibility of an evaluation result is improved.

Description

technical field [0001] The invention relates to the field of mechanical automation testing, in particular to a method and system for evaluating the response of a physical robot based on deep learning. Background technique [0002] The response speed of the physical robot is a very important indicator of the robot, which directly affects the flexibility of the robot. At present, there are various evaluation standards for robots, but they are only tested for the overall robot response speed, and the test method is very cumbersome and the accuracy is not high. In the process of developing the robot, it is necessary to evaluate the response speed of each execution link of the robot, so that the overall response speed of the robot can reach the standard, and even reach the peak of the industry. [0003] Deep learning (DL, Deep Learning) is a new research direction in the field of machine learning (ML, Machine Learning). It is introduced into machine learning to make it closer to...

AI Technical Summary

Problems solved by technology

[0004] There are the following disadvantages in the test of passionate human response in the prior art: 1. Most of the physical robot response tests in the industry are to wait for the robot to be assembled, and use manual and stopwatch methods to test its overall response speed. This test method is cumbersome and tedious. Low test efficiency

[0005] 2. The physical robot is composed of multiple modules such as mechanical arm, voice dialogue, face recognition, video image, etc., and there are also multiple components in each module, such as voice dialogue, which involves the microphone's sound pickup ability and voice recognition Wait, if you only test on the whole machine, there will be problems that cannot be accurately located. Which node is slow in responding?

[0006] 3. The physical robot is composed of software and hardware. Each device will be somewhat different, and the use environment, such as the network, will also affect the response speed of the robot. It is very important to monitor and evaluate each physical robot. Necessary, if the current manual technology is used for testing, the corresponding effect cannot be achieved, and the labor cost is high. It is impossible to conduct multiple on-site tests for some robots in operation

[0007] 4. Every time a physical robot is added or a new capability is added, it needs to be evaluated. This is a huge workload, and manual evaluation is also very laborious, and manual capabilities cannot be achieved.

Images