The Design of Battle Field Robot

Transcription

1 The Design of Battle Field Robot D.Bubesh Kumar Mechanical Engineering Department Aarupadaiveedu Institute of Technology Vinayaka Mission University Kanchipuram,Tamilnadu,India Abstract -The battle field robot is the robot of the future warfare. The future war will be fought with nuclear, chemical and Biological weapons,where it is difficult for human soldiers to fight the enemies.the battle field robot consists of CCD cameras for eye, servomotors for locomotion, Fuzzy neural network for the brain of the robot. The fuzzy rules are developed and stored in the microprocessor of the robot, and from the microprocessor the fuzzy rules are sent to the neural network,separate servo motors are placed at the locomotion points of the robot to achieve moments similar to the humans.the fuzzy neural network is trained by the three dimensional multilayered fuzzy neural network, the robot gets signal from the main computers which is placed in the military base.the kinematics of essential movements of the robot are derived with the free body diagram. The training of the neural network is also explained in this paper. Keywords:Neural Network, Fuzzy, Humanoid 1 Introduction chips The battle field Robot is designed to carry a payload of 100kg of arms and ammunitions rocket launchers and laser guns.its movements are similar to the bipod robot. It has arms and legs similar to the humanoid. The arms can swivel for 360 degree. The head of the robot can rotate 360 degree, to enable to identify the object in 3D. The robot has a global positioning system which guides the robot to the required destination.it also has a control box linked to the base control room from which tasks are given in the form of instructions, from the given instructions the battle field robot makes is own decisions and also using the neural chip to achieve its objectives. 3 The essential movements of the robot are designed to qualify for the battle field 1. Walking. Jumping 3. Running 4. Combat 5. Shooting 6. Stealth movements 7. Staircase climbing 8. Squatting In today s modern warfare, many countries are in possession of nuclear, chemical and biological weapons, the soldiers have to work in a dangerous environment. So battle field Robots play a vital role in the battle field. Battle field robots are used in direct combat and in guarding the nuclear power plants and military installations. The robots are designed similar to the Humanoids. This paper deals with locomotion of the robots, the kinematics of the linkages and equations are formed using Euler angles to design the Robot to have movements similar to human movements. Construction of battle field robot The robot consists of power source usually from battery. CCD cameras are used as the eyes of the robot, the control or the brain of the robot, neural network chip are used. The servo motors which actuates the arms and legs of the robot are controlled by the neural figure Squatting of Battle Field Robot

2 Figure 1 Battle Field Robot muscle (servo 7)connects the foot to the thigh (servo 6) the quad muscle connects the thigh and the leg through a frictionless pulley mechanism representing the patella joint. The joints at H,K and A are hinge joints. The tension in cord (hamstring, calf) and pulley (quads) as a function of the angle the leg makes with the horizontal plane (ө)[1]. For simplicity we assume symmetry with respect to the horizontal plane passing through the knee joint.let P denote the force transmitted at the hip joint to each leg, the force P is the equal to half the weight of the upper body plus the weight of the squat plus the half the weight of the ammunition carried by the battle field robot as shown in figure 3. Consider the free body diagram of the foot shown in fig 4a. The movement produced by the force P about A should be equal to the movement produced by the calf muscle at A. ( servo 7) as shown in figure 4b. figure 3 FBD of lower portion of battle field robot 4 The squatting performed by a battle field robot The movement is slow enough to assume static equilibrium a link system is shown in the fig to represent squat. The beam representing hip is connected to the rod representing the upper leg at the hip joint. A tension carrying cord representing the calf bsinϕf F = F ( L cosθ c) P d F h L sinϕ = ( L+ h)sinθ h c p C + cp= 0= cp /( bsinϕ) L cosϕ = b+ ( L h)cosθ sinϕ+ d F tanϕ = ( L+ h)sinθ /[ b+ ( L h)cosθ The ø is expressed in the form of ө k h q q = 0 (1) () (3) (4) (5) (6)

3 figure 4a FBD of foot figure 4c FBD Knee joint c figure 4 d FBD of patella figure 4 b FBD of calf 5 The determination of movement Arm the compressive force acting on the quads and the knee joint can be found by considering the static equilibrium of K as shown in figure 4d. P q F = F cosα (9) d k of the quad muscle group,the geometry associated with quad muscle is shown in figure 4c. d s = ( L / 3) q = u sinα + u (L / 3) u cos( π θ ) (7) (8)

4 6 The vertical jump Vertical jump performed by battle field robot to reach higher places to leap over obstacles. The propulsive stage of the vertical jump, the battle field root. The segments of the battle field robot change to lift the upper body 70% of the body weight vertically, The mechanic of vertical jumping can be captured with reasonable accuracy by using a four segment model, fig 5a, consists of foot shank, thigh and the upper body. The geometry of the assumed structure dictates that the spatial position of the point mass M at any time t is given by the equation. The figure 5b shows the FBD jumping battle field robot. r= Lsinθe v= Lcosθ ( dθ e a= [ L sinθ( dθ + Lcosθ ( d θ / dt )] e P Mge = M[ L sinθ ( dθ + Lcosθ ( d θ / dt )] e P= 0 g= Lsinθ ( dθ Lcosθ ( d θ / dt ) Aftertakeoff a= -ge v= (1.65 gt) e r= (1.65t 0.5gt ) e (10) (11) (1) (13) (14) (15) figure 5b FBD jumping battle field robot figure 5a jumping robot

5 figure 7 Fuzzy Artificial Neural Network 7 Design of neural network In this paper the humanoid robot control has been designed using fuzzy artificial neural network.fuzzy neural network fig 7 combine the best of both worlds. They consider fuzzy inputs and at the same time, they are capable of learning.here the forward network is fuzzy and the feed back is neural network[]. The neural network fig 8 receives the inputs and the actual outputs, it creates a new classifications and input output associations and it generates new rules. It updates the forward network with new rules. more layers will be involved in the learning process, to achieve more human like thinking and movements. 9 References [1] Human Body Dynamics :Classical Mechanics and Human Movement,Aydin Tozeren, Springer,000 [] Understanding Neural Networks ad Fuzzy Logic,Stamatoos V. Karatalopoulous,IEEE Press,004 8 Conclusions The battle field robot involves more research in the flied of neural network and kinematics of human joints. In this paper one layered network was used to train the robot in the initial phase and in the subsequent phases

6