Robotic Tool Collision Detection with Adaptive Motor Current Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Self-propelled robotic tools, such as robotic lawnmowers, face challenges in reliably detecting collisions due to varying environmental conditions and tool aging, leading to false or missed collision detections.
Innovation Solution
The method involves updating prediction algorithm parameters based on current errors exceeding thresholds, allowing the algorithm to adapt to specific circumstances like thick grass or tool aging, using sensors for ground and weather adjustments, and implementing a counter-based updating mechanism to refine predictions over time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a fixed prediction algorithm is used for collision detection, then the algorithm is simple to implement, but it produces false detections when environmental conditions change or the tool ages
Solution Approach 1:
The prediction algorithm transitions from a static, fixed form to a dynamic, adaptive form. The algorithm parameters are continuously updated based on actual operational data, allowing the system to adapt to changing environmental conditions and tool aging while maintaining collision detection reliability.
Solution Approach 2:
A feedback mechanism is introduced where the actual motor current measurements are compared with predicted values, and the prediction algorithm parameters are updated based on the errors. This closed-loop feedback system enables the algorithm to learn from operational experience and improve its accuracy over time.
2Adaptability or versatility
If the prediction algorithm parameters are frequently updated, then the algorithm adapts quickly to changing conditions, but the system becomes more complex and computationally intensive
Solution Approach 1:
The parameter update process is implemented periodically rather than continuously. The first control loop runs at high frequency for collision detection, while the second control loop updates parameters at a lower frequency based on accumulated error statistics, balancing adaptability with system simplicity.
Solution Approach 2:
The system performs preliminary error accumulation and statistical analysis before updating parameters. By maintaining error counters and evaluating thresholds before parameter updates, the system prepares in advance to avoid unnecessary updates and reduce computational overhead.
Data Source
AI summary
The present disclosure relates to a self-propelled robotic work tool (1), e.g. an automatic robotic lawn mower, and a corresponding method. The robotic tool comprises an inertia measurement unit (IMU 15) which generally obtains (25) measured IMU parameters regarding the robotic working tool's movement. A prediction algorithm (17) predicts (27,36) required motor currents for driving the robotic work tool's wheels (5) based on the measured IMU parameters. The predicted motor current is compared (29,37) to the actual current used and the difference constitutes an error (19), which is used in a collision detection unit (21). If the collision detection unit (21) senses that the actually used motor current is much higher than the predicted current, a collision may be indicated (31). The prediction algorithm is repeatedly updated based on the error (19) by incrementing or decrementing an error category counter (error cat, 41,45) if the error (19) is above or bellow a first or second threshold (39,43), and increasing or decreasing a prediction algorithm setting (49,53), e.g. a motor current offset term i offset—if the error category counter (error cat, 41,45) is above or bellow a third or fourth threshold (47,51). This allows the prediction algorithm to adapt to circumstances where the robotic tool is used. For instance, if a lawn mower operates in thick grass, the prediction algorithm can be adapted not to detect false collisions due to increased motor current values.