According to the requirements of the live-virtual-constructive(LVC)tactical confrontation(TC)on the virtual entity(VE)decision model of graded combat capability,diversified actions,real-time decision-making,and genera...According to the requirements of the live-virtual-constructive(LVC)tactical confrontation(TC)on the virtual entity(VE)decision model of graded combat capability,diversified actions,real-time decision-making,and generalization for the enemy,the confrontation process is modeled as a zero-sum stochastic game(ZSG).By introducing the theory of dynamic relative power potential field,the problem of reward sparsity in the model can be solved.By reward shaping,the problem of credit assignment between agents can be solved.Based on the idea of meta-learning,an extensible multi-agent deep reinforcement learning(EMADRL)framework and solving method is proposed to improve the effectiveness and efficiency of model solving.Experiments show that the model meets the requirements well and the algorithm learning efficiency is high.展开更多
In order to ensure that the system has the advantage of light weight and vibration absorption, the steel rope is used as a flexible transmission part. A flexible drive unit(FDU) is developed, whose features are guided...In order to ensure that the system has the advantage of light weight and vibration absorption, the steel rope is used as a flexible transmission part. A flexible drive unit(FDU) is developed, whose features are guided by steel rope, increasing force by the movable pulley group, modular, convenient and flexible. Dynamics model for controller is deduced based on the constitutive equation of viscoelasticity. Controller is designed for position control and is based on the viscoelasticity dynamics model compensation control strategy proposed. The control system is based on the TURBO PMAC multi-axis motion control card.Prototype loading experiments and velocity experiments results show that the FDU can reach 2 Hz with no load and the max speed of 30(°)/s. The FDU has the capability of the load torque 11.2 N·m and the speed of 24(°)/s simultaneously, and the frequency response is 1.3 Hz. The FDU can be used to be the pitch joint of hip for biped robot whose walking speed is 0.144 km/h theoretically.展开更多
单元测试用于检验软件单一模块的功能是否正确,是软件开发过程中的重要步骤,可以及时发现代码中的缺陷,提升软件的质量和可信度.由于手动编写单元测试费时费力,经常遗漏覆盖重要的代码逻辑.为此,研究者提出单元测试用例自动生成技术.近...单元测试用于检验软件单一模块的功能是否正确,是软件开发过程中的重要步骤,可以及时发现代码中的缺陷,提升软件的质量和可信度.由于手动编写单元测试费时费力,经常遗漏覆盖重要的代码逻辑.为此,研究者提出单元测试用例自动生成技术.近来,预训练大语言模型(large language models,LLM)已经广泛应用于代码生成相关任务.然而,当前在重要的系统级编程语言C上,还没有相关工作.为了填补这一空白,本文面向C程序设计并实现了基于LLM的单元测试用例生成方法LLM4CUTCG.该方法结合LLM多智能体交互和程序分析技术,客服了LLM内在问题.为了验证方法效果,收集了125个C语言目标程序,并针对这些程序生成测试用例.实验结果表明,LLM4CUTCG生成的测试行覆盖率为91.71%,测试预言正确率为50.05%.其覆盖率优于传统方法符号执行.展开更多
基金supported by the Military Scentific Research Project(41405030302,41401020301).
文摘According to the requirements of the live-virtual-constructive(LVC)tactical confrontation(TC)on the virtual entity(VE)decision model of graded combat capability,diversified actions,real-time decision-making,and generalization for the enemy,the confrontation process is modeled as a zero-sum stochastic game(ZSG).By introducing the theory of dynamic relative power potential field,the problem of reward sparsity in the model can be solved.By reward shaping,the problem of credit assignment between agents can be solved.Based on the idea of meta-learning,an extensible multi-agent deep reinforcement learning(EMADRL)framework and solving method is proposed to improve the effectiveness and efficiency of model solving.Experiments show that the model meets the requirements well and the algorithm learning efficiency is high.
基金Project(2006AA04Z201,2012AA041601)supported by the National High-Tech Research and Development Program of China
文摘In order to ensure that the system has the advantage of light weight and vibration absorption, the steel rope is used as a flexible transmission part. A flexible drive unit(FDU) is developed, whose features are guided by steel rope, increasing force by the movable pulley group, modular, convenient and flexible. Dynamics model for controller is deduced based on the constitutive equation of viscoelasticity. Controller is designed for position control and is based on the viscoelasticity dynamics model compensation control strategy proposed. The control system is based on the TURBO PMAC multi-axis motion control card.Prototype loading experiments and velocity experiments results show that the FDU can reach 2 Hz with no load and the max speed of 30(°)/s. The FDU has the capability of the load torque 11.2 N·m and the speed of 24(°)/s simultaneously, and the frequency response is 1.3 Hz. The FDU can be used to be the pitch joint of hip for biped robot whose walking speed is 0.144 km/h theoretically.
文摘单元测试用于检验软件单一模块的功能是否正确,是软件开发过程中的重要步骤,可以及时发现代码中的缺陷,提升软件的质量和可信度.由于手动编写单元测试费时费力,经常遗漏覆盖重要的代码逻辑.为此,研究者提出单元测试用例自动生成技术.近来,预训练大语言模型(large language models,LLM)已经广泛应用于代码生成相关任务.然而,当前在重要的系统级编程语言C上,还没有相关工作.为了填补这一空白,本文面向C程序设计并实现了基于LLM的单元测试用例生成方法LLM4CUTCG.该方法结合LLM多智能体交互和程序分析技术,客服了LLM内在问题.为了验证方法效果,收集了125个C语言目标程序,并针对这些程序生成测试用例.实验结果表明,LLM4CUTCG生成的测试行覆盖率为91.71%,测试预言正确率为50.05%.其覆盖率优于传统方法符号执行.
