Robot Controller Pseudo Inverse Matrix Determination

The patent applies dynamics by making the increment of the adjustment parameter k variable rather than fixed. The increment is dynamically adjusted based on the current value of k and the determinant value, allowing the determination process to adapt to different stages of calculation. This dynamic adjustment ensures that k changes smoothly and continuously, preventing discontinuous jumps in manipulation amount while maintaining the exploratory determination approach.

The patent changes the parameter of increment from a fixed constant to a variable that depends on the current state of calculation. Specifically, the increment is set as a function of the current adjustment parameter k and the determinant value, allowing the system to automatically adjust the step size during the exploratory process. This parameter change enables both simple operation and smooth motion by adapting the increment to the calculation context.

The patent makes the increment dynamic by setting it as a function of the current adjustment parameter k and the determinant value. When k is small, the increment is larger to speed up the determination process. When k approaches the optimal value or the determinant becomes sufficiently large, the increment automatically decreases to ensure smooth convergence. This dynamic behavior allows fast determination while preventing frequent large changes in k that would cause discontinuous manipulation amounts.

The patent implements feedback by using the determinant value and current k value to adjust the increment for the next step. The system continuously monitors the determinant value and adjusts the increment accordingly - reducing the increment when the determinant is already sufficiently large or when k is close to the optimal value. This feedback mechanism ensures that the determination process remains fast while maintaining continuity in manipulation amount by preventing excessive changes in k.

The patent applies dynamics by making the increment variable rather than uniformly small. The increment starts larger when k is small, allowing rapid initial progress. As k increases and approaches the optimal value, the increment automatically decreases to ensure smooth convergence. This dynamic adjustment maintains manipulation amount smoothness while minimizing control cycle duration by avoiding unnecessarily small increments during early stages of determination.

The patent changes the increment parameter from a fixed small value to a variable that adapts to the calculation progress. The increment is set as a function of current k and determinant value, allowing it to be large when appropriate (reducing control cycle time) and small when necessary (maintaining smoothness). This parameter change resolves the contradiction by enabling the system to use small increments only when needed for smooth convergence, rather than using small increments throughout the entire process.