Manipulator Control State Machine for Position-Force Transitions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Classic control state machines for manipulators, such as industrial robots, face challenges in seamlessly blending between position- and force-controlled sections, leading to cycle time losses and difficulties in synchronizing filters and observers, due to their deterministic and rigid implementation.
Innovation Solution
Implementing a state machine architecture where commands pass through pre-run, active, and post-run states, allowing control components to settle and synchronize in advance, enabling parallel processing and maintaining a well-defined state for orderly transitions, with additional states for safety and error handling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If classic control state machines run through individual states in a determined sequence with stationary exact stops, then the control structure is simple and deterministic, but cycle time losses occur and transitions between position- and force-controlled sections are difficult
Solution Approach 1:
The control method segments the execution of a command into multiple distinct phases: a pre-run phase for initialization and filtering, an active phase for execution, and a post-run phase for finalization. This segmentation allows parallel processing of multiple commands at different phases, reducing cycle time losses while maintaining a manageable control structure through structured state management.
Solution Approach 2:
The pre-run phase performs preliminary actions by initializing control components and allowing filters and observers to settle before the command enters the active phase. This preliminary preparation enables smoother transitions between control modes and eliminates waiting time during state transitions, directly addressing the cycle time loss problem.
2Measurement precision
If filters, observers and parameter identifications need lead time to settle down, then measurement precision is improved, but cycle time losses occur due to waiting for settlement
Solution Approach 1:
The pre-run phase is specifically designed to provide the necessary lead time for filters, observers, and parameter identifications to settle down before the command becomes active. By performing this settlement process in advance during the pre-run phase, the system ensures measurement precision is achieved without causing cycle time losses during the active execution phase.
Solution Approach 2:
The control system dynamically transitions commands between pre-run and active phases, allowing flexible timing for filter settlement. Multiple commands can be at different stages simultaneously, with some in pre-run allowing filters to settle while others are actively executing, thus optimizing both measurement precision and cycle time utilization.
3Adaptability or versatility
If a rigid deterministic implementation is used, then the control structure is simple, but adaptability for connecting external processes or modifying control is difficult
Solution Approach 1:
The state machine implementation is made dynamic by allowing commands to transition between pre-run and active phases based on readiness conditions rather than fixed deterministic sequencing. This dynamic approach enables external processes to trigger command activation and allows flexible modification of control flow while maintaining a relatively simple state machine structure through the use of well-defined state transitions.
Solution Approach 2:
The system allows parameter changes and external process connections by modifying the control state from pre-run to active based on external triggers or conditions. This parameter change approach enables adaptability without requiring fundamental changes to the control structure, as the same state machine framework handles both deterministic and externally-triggered command execution.
Data Source
Figure 1
AI summary
The method involves processing a set of control commands e.g. hold pose using model-based regulation with observer control command (B), impose a desired force using force regulation control command (F) and travel to desired pose under position regulation control command (P). The commands are executed in a state machine in an active state (A). Each of the commands is executed in the state machine in a preliminary state that is placed before the active state and/or a post-operational state that is placed after the active state, and the set of commands are processed simultaneously. Independent claims are also included for the following: (1) a control device for controlling a manipulator comprising a device for executing a method for controlling a manipulator (2) a computer program for executing a method for controlling a manipulator (3) a computer program product comprising a program code for executing a method for controlling a manipulator.