Method of robot dynamic motion planning and control

A method for dynamic motion planning for a robot, said method executed on a computing device and comprising:

• computing a design robot motion based on a target location of a robot tool center point,
• computing a commanded robot motion for a current control cycle of calculated motion for the robot based on the design robot motion and obstacle data received from sensors, where the commanded robot motion is provided as feedback for computing the design robot motion for a next control cycle, and where computing a commanded robot motion includes an optimization computation having a multi-term objective function with coefficients being adjustable at each control cycle, position, velocity and acceleration limitations for one or more robot joints defined as inequality constraints, and a collision avoidance constraint;
• and
• providing the commanded robot motion to a controller module which sends signals control of joint motions of the robot,
• where the design robot motion is a tool center point motion in Cartesian space, and the commanded robot motion includes robot joint rotational motions for all joints in the robot, and the robot has at least one redundant degree of freedom such that a number of degrees of freedom of the robot joint rotational motions exceeds a number of degrees of freedom of the tool center point motion.

A dynamic motion planning system for an industrial robot, said system comprising:

• a perception module including at least one sensor or camera configured to detect obstacles in a workspace of the robot;
• and
• a computing device having a processor and memory, configured with:
• a planning module programmed to compute a design robot motion based on a target location of a tool center point on a tool attached to the robot;
• and
• a motion optimization module receiving obstacle data from the perception module, said motion optimization module being programmed to compute a commanded robot motion for a current control cycle of calculated motion for the robot based on the design robot motion and the obstacle data, where the commanded robot motion is provided as feedback to the planning module for computing the design robot motion for a next control cycle, and where the commanded robot motion is computed using an iterative optimization computation having a variably-weighted multi-term objective function, one or more robot structural limitation constraints and a collision avoidance constraint, where the multi-term objective function includes a path deviation minimization term, a joint velocity regularization term and a robot pose regularization term, and the objective function seeks a minimum of a sum of the terms,
• and where the commanded robot motion is provided to a controller module which sends signals executing control of joint motions of the robot.

A method for dynamic motion planning for a robot, said method executed on a computing device and comprising:

• computing a design robot motion based on a target location of a robot tool center point;
• computing a commanded robot motion for a current control cycle of calculated motion for the robot based on the design robot motion and obstacle data received from sensors, where the commanded robot motion is provided as feedback for computing the design robot motion for a next control cycle, and where computing a commanded robot motion includes an optimization computation having a multi-term objective function, one or more robot structural limitation constraints and a collision avoidance constraint, where the collision avoidance constraint is an inequality constraint computed from a safety function, where the collision avoidance constraint is that a rate of change of the safety function must be greater than or equal to the negative of the safety function multiplied by a coefficient;
• and
• providing the commanded robot motion to a controller module which sends signals executing control of joint motions of the robot.