Robot Actuator and Control Co-Design for Multi-Task Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robotic manipulator designs face challenges in co-designing motion and motors for multiple tasks due to heavy computational burden and lack of flexibility, often resulting in suboptimal performance and limited design freedom.
Innovation Solution
A sequential co-design strategy that incorporates probabilistic approaches for jointly computing trajectories and motor designs, using magnetic equivalent circuit modeling and a differentiable simulator to optimize motor geometry and control systems for specific applications, allowing for gradient-based optimization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional separate design of mechanical subsystem and control subsystem is used, then design simplicity is maintained, but system performance and flexibility are insufficient
Solution Approach 1:
The patent merges the mechanical subsystem design and control subsystem design into a unified co-design framework. The sequential co-design strategy simultaneously optimizes motor parameters (structural parameters) and control policy parameters, allowing the two previously separate design processes to interact and inform each other, thereby achieving improved system flexibility without excessive design complexity
Solution Approach 2:
The patent introduces dynamic adaptability through the differentiable simulator that allows online or near-online adjustment of control parameters based on actual system performance. The control policy is designed to be adaptable rather than fixed, enabling the system to adjust to different tasks and conditions while maintaining a relatively simple mechanical structure
2Productivity
If full co-design optimization for multiple tasks is performed, then system performance is improved, but computational burden becomes prohibitive
Solution Approach 1:
The patent segments the complex co-design problem into a sequential process with distinct phases: first optimizing motor parameters, then optimizing control policy parameters, and finally integrating them. This segmentation allows each sub-problem to be solved more efficiently than attempting to solve the entire problem simultaneously, reducing the overall computational burden while still achieving multi-task optimization
Solution Approach 2:
The patent performs preliminary optimization of motor structural parameters before proceeding to control policy optimization. This preliminary action establishes a solid foundation that simplifies subsequent control design and reduces the computational search space for the second phase, thereby reducing total computational time while maintaining high system performance
3Adaptability or versatility
If fixed electromechanical design is used, then manufacturing simplicity is maintained, but optimization flexibility for different tasks is limited
Solution Approach 1:
The patent implements dynamic adaptability primarily in the control software layer rather than requiring physical reconfiguration of the mechanical system. The differentiable simulator and optimized control policies enable the system to adapt to different tasks through software adjustments, maintaining manufacturing simplicity while achieving high task optimization flexibility
Solution Approach 2:
The patent optimizes specific motor parameters (such as inertia, torque constants, and gear ratios) and control parameters (such as PID gains and trajectory parameters) to achieve better performance across multiple tasks. These parameter optimizations are carefully selected to provide substantial performance improvements while requiring minimal manufacturing changes, thus balancing manufacturing ease with task flexibility
Data Source
AI summary
An engineering system comprises a memory having instructions stored thereon and at least one processor configured to execute the instructions to cause the system to collect a plurality of tasks for a manipulator actuated by a motor. Structural parameters of the motor, a plurality of reference trajectories of the motor for actuating the manipulator to perform the plurality of tasks, and parameters of a feedback control policy for the manipulator are jointly determined to increase overlap between a probability distribution of values of operational data of the motor operating according to different real trajectories from a plurality of real trajectories and an efficiency map of the motor defined in a domain of the operational data of the motor. The structural parameters of the motor, the plurality of reference trajectories, and the feedback control policy are output for performing the plurality of tasks.