Multi-constraint handling and a mixed integer predictive controller for space robots with obstacle avoidance

For the issue that obstacles need to be avoided in many space robots tasks, this paper develops a mixed integer predictive controller for space robots avoiding obstacles when performing tasks. Firstly, an improved obstacle avoidance constraint is formulated based on propositional logic. Then, in the frame of Model Predictive Control (MPC) method, tracking errors and fuel consumptions of all manipulator joints are involved in the cost function, and three types of constraints, joint input and output limits, as well as the developed obstacle avoidance constraint, compose the inequality constraints. Furthermore, the constraint priority is established based on propositional logic, guaranteeing the problem could be solved under the satisfaction of maximum number of the constraints. Simulation results illustrate the improved obstacle avoidance constraint based on propositional logic could be better for particle following the reference trajectory than the traditional one. And the mixed integer predictive controller effectively ensures avoiding obstacles during space robots performing the tasks.