Walking Control Using Ballistic Hip and Heel Impulse
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robot control systems for walking on limbs struggle to maintain efficient movement over long periods due to insufficient motorization and reliance on external energy sources, particularly in environments with varying terrain, and are computationally demanding, limiting their speed and accuracy.
Innovation Solution
A method that eliminates the need for Zero Moment Point (ZMP) by using ballistic hip movements and kinetic energy supplied through heel activation, allowing for more effective modeling of human-like walking by computing new bearing points based on the robot's center of mass speed and controlling limb elongation through ankle, knee, or cylinder impulses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the Zero Moment Point (ZMP) concept is used to control robot walking, then dynamic balance conditions are satisfied, but the computational power required becomes extremely demanding and precise knowledge of ground topography is needed
Solution Approach 1:
The patent extracts and eliminates the ZMP concept from the control system, replacing it with a simplified approach that uses the robot's center of mass position and velocity directly to determine foot placement, thereby removing the computationally intensive ZMP calculations while maintaining balance control
Solution Approach 2:
The control method uses the robot's own motion state (center of mass position and velocity) to automatically determine the next foot placement position without requiring external ground topology information or complex iterative computations, enabling the system to serve itself with minimal computational overhead
2Reliability
If the ZMP concept is used to control robot walking, then dynamic balance is achieved, but the system requires precise knowledge of ground topography which reduces adaptability
Solution Approach 1:
The control method uses the robot's own motion state (center of mass position and velocity) to automatically determine the next foot placement position without requiring external ground topology information or complex iterative computations, enabling the system to serve itself with minimal computational overhead
3Speed
If hip motorization is used to provide kinetic energy for walking, then forward progression is achieved, but the energy is insufficient to compensate for friction losses over long periods
Solution Approach 1:
The patent implements periodic heel activation that occurs at specific phases of the walking cycle, providing intermittent energy boosts that compensate for friction losses accumulated during the ballistic hip movement phases, enabling sustained walking over long periods
Solution Approach 2:
The control system combines two different movement mechanisms - ballistic hip movements for efficient forward progression and periodic heel activation for energy replenishment - creating a composite walking gait that leverages the advantages of both approaches to achieve both speed and duration
4Productivity
If heel activation is used to supply kinetic energy, then walking efficiency is improved, but the control system becomes more complex
Solution Approach 1:
The patent implements periodic heel activation that occurs at specific phases of the walking cycle, providing intermittent energy boosts that compensate for friction losses accumulated during the ballistic hip movement phases, enabling sustained walking over long periods
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enables robots to move faster and more efficiently on uneven surfaces without requiring precise knowledge of the terrain, replicating human-like walking patterns and reducing computational demands, allowing for speeds of up to 5 Km/h with improved stability and realism.
Implementation Method 1
a first step of movement of each of the hips over a substantially spherical trajectory centered on the end bearing on the ground of the limb attached to said hip
Implementation Method 2
combining a model of the walk driven by the ballistic displacements of the hips with a supply of kinetic energy to the robot by the activation of the heel
Implementation Method 3
the elongation between the end of the limb bearing on the ground and said hip is augmented by a value such that the segment joining the position of said hip at the end of a spherical trajectory to the position of the second hip at the start of the next spherical trajectory is substantially parallel to the progression surface
Data Source
AI summary
A computer program and a system for controlling walking of a mobile robot, notably a humanoid robot moving on two legs. Conventionally, control was guided by driving a zero moment point. Such driving was performed within a fixed coordinate system connected to a progression surface and assumed knowledge of the characteristics of said surface and the creation of a provisional trajectory. Such driving encountered significant limitations due to the nature of the progression surfaces on which walking can effectively be controlled and an obligation to have a perfect knowledge of their geometry; and also in respect to the necessary computing power, and the appearance of the walk which bore little resemblance to an actual human walk. The invention overcomes such limitations by providing a walk which includes a pseudo-free or ballistic phase, an impulse phase imparted by the heel of the robot, and a landing phase.