<p>In order to plan dual-arms of wheeled mobile humanoid robot while it performs the end-effector tasks, a hybrid constrained motion planning (HCMP) scheme is proposed and analyzed. In traditional schemes, obstacle avoidance between a single manipulator and an obstacle is common. Different from traditional schemes, the proposed HCMP scheme can not only avoid the collision between dual-arms of humanoid robot, but also allow end-effector to keep a specific angle. To do so, firstly, the creation of the inverse kinematic model of the robot is achieved. Secondly, the mutual avoidance constraints, improved two-element posture maintaining constraints and joint physical constraints are designed. Thirdly, the kinematic model and multiple constraints are transformed to a HCMP scheme via quadratic programming. Fourthly, in order to obtain the optimal solution under the HCMP scheme, a linear variable inequality primal dual neural network is used. Finally, the simulation verification shows that the proposed HCMP scheme can achieve the motion planning of the wheeled mobile humanoid robot.</p>