From 1605d304fb7b94c2a577a7f95a4667057f9c44e1 Mon Sep 17 00:00:00 2001 From: Sai Raghava Reddy Ganapaa Date: Tue, 18 Jun 2019 11:59:24 +0200 Subject: [PATCH] draft for the entire code using the ros-publisher --- spider_control/control.py | 107 ++++++++++++++++++++++++++++++++++++-- 1 file changed, 103 insertions(+), 4 deletions(-) diff --git a/spider_control/control.py b/spider_control/control.py index a189eb9..87681b3 100644 --- a/spider_control/control.py +++ b/spider_control/control.py @@ -127,7 +127,7 @@ def call_return_joint_states(joint_names): print"joint %s not found"%joint_name return (resp.position, resp.velocity, resp.effort) -def talker(jointname, position): +def talker(jointname , position): pub1 = rospy.Publisher('joint_states', JointState, queue_size = 10) rospy.init_node('joint_state_publisher') rate = rospy.Rate(10) @@ -463,7 +463,7 @@ if __name__=="__main__": if ( run_method == 'train'): training_time = raw_input("please enter how long you intend to train the algorithm") run_time = training_time - model_trained = raw_input("which model do you want to train ?, please enter model_number for straight terrain select 1, for accute angled terrain select 2, for obtuse angled terrain select 3") + model_trained = raw_input("which model do you want to train ?, please enter model_number for straight terrain select 1, for accute angled terrain select 2, for obtuse angled terrain select 3, for up/down terrain select 4") output = one_hot_encoding(run_time, model_trained) pref_out = tf.placeholder(tf.float32, shape=(4, 1)) with tf.Session as sess1: @@ -595,7 +595,6 @@ if __name__=="__main__": input_layerout.append(tf.multiply(output_layerhid, model_4.weights_o[16:24])) # element wise multiplication model_4.get_neuron_value(model_4.neuron_o, input_layerout) output_layerout = model_4.output_function(model_4.neuron_o) # output of the neural network for model 1 - x_left_com = Leg_attribute.x_left_com(Leg_attribute.m_l, g, leg_1.r_1, leg_1.l_1, joint_angle_leg_1[0], joint_angle_leg_1[1], joint_angle_leg_1[2], joint_angle_leg_2[0], joint_angle_leg_2[1], joint_angle_leg_2[2], joint_angle_leg_3[0], @@ -875,7 +874,107 @@ if __name__=="__main__": 'finding the softmax output of the neurons' softmax_output = np.array(4) softmax_output = knowledge_transfer.out_softmax(knowledge_transfer.neuron_out)#this gives the softmax output and stores it in the newly created array - + model = max(softmax_output) + if (softmax_output[0] == model): + talker('joint_1_1', output_layerout[0]) + talker('joint_1_2', output_layerout[1]) + talker('joint_1_3', output_layerout[2]) + talker('joint_4_1', output_layerout[9]) + talker('joint_4_2', output_layerout[10]) + talker('joint_4_3', output_layerout[11]) + talker('joint_5_1', output_layerout[12]) + talker('joint_5_2', output_layerout[13]) + talker('joint_5_3', output_layerout[14]) + talker('joint_8_1', output_layerout[21]) + talker('joint_8_2', output_layerout[22]) + talker('joint_8_3', output_layerout[23]) + talker('joint_2_1', output_layerout[3]) + talker('joint_2_2', output_layerout[4]) + talker('joint_2_3', output_layerout[5]) + talker('joint_3_1', output_layerout[6]) + talker('joint_3_2', output_layerout[7]) + talker('joint_3_3', output_layerout[8]) + talker('joint_6_1', output_layerout[15]) + talker('joint_6_2', output_layerout[16]) + talker('joint_6_3', output_layerout[17]) + talker('joint_7_1', output_layerout[18]) + talker('joint_7_2', output_layerout[19]) + talker('joint_7_3', output_layerout[20]) + elif (softmax_output[1] == model): + talker('joint_7_1', output_layerout[18]) + talker('joint_7_2', output_layerout[19]) + talker('joint_7_3', output_layerout[20]) + talker('joint_2_1', output_layerout[3]) + talker('joint_2_2', output_layerout[4]) + talker('joint_2_3', output_layerout[5]) + talker('joint_1_1', output_layerout[0]) + talker('joint_1_2', output_layerout[1]) + talker('joint_1_3', output_layerout[2]) + talker('joint_4_1', output_layerout[9]) + talker('joint_4_2', output_layerout[10]) + talker('joint_4_3', output_layerout[11]) + talker('joint_3_1', output_layerout[6]) + talker('joint_3_2', output_layerout[7]) + talker('joint_3_3', output_layerout[8]) + talker('joint_6_1', output_layerout[15]) + talker('joint_6_2', output_layerout[16]) + talker('joint_6_3', output_layerout[17]) + talker('joint_5_1', output_layerout[12]) + talker('joint_5_2', output_layerout[13]) + talker('joint_5_3', output_layerout[14]) + talker('joint_8_1', output_layerout[21]) + talker('joint_8_2', output_layerout[22]) + talker('joint_8_3', output_layerout[23]) + elif (softmax_output[2] == model): + talker('joint_1_1', output_layerout[0]) + talker('joint_1_2', output_layerout[1]) + talker('joint_1_3', output_layerout[2]) + talker('joint_8_1', output_layerout[21]) + talker('joint_8_2', output_layerout[22]) + talker('joint_8_3', output_layerout[23]) + talker('joint_3_1', output_layerout[6]) + talker('joint_3_2', output_layerout[7]) + talker('joint_3_3', output_layerout[8]) + talker('joint_2_1', output_layerout[3]) + talker('joint_2_2', output_layerout[4]) + talker('joint_2_3', output_layerout[5]) + talker('joint_5_1', output_layerout[12]) + talker('joint_5_2', output_layerout[13]) + talker('joint_5_3', output_layerout[14]) + talker('joint_4_1', output_layerout[9]) + talker('joint_4_2', output_layerout[10]) + talker('joint_4_3', output_layerout[11]) + talker('joint_7_1', output_layerout[18]) + talker('joint_7_2', output_layerout[19]) + talker('joint_7_3', output_layerout[20]) + talker('joint_6_1', output_layerout[15]) + talker('joint_6_2', output_layerout[16]) + talker('joint_6_3', output_layerout[17]) + elif (softmax_output[3] == model): + talker('joint_1_1', output_layerout[0]) + talker('joint_1_2', output_layerout[1]) + talker('joint_1_3', output_layerout[2]) + talker('joint_3_1', output_layerout[6]) + talker('joint_3_2', output_layerout[7]) + talker('joint_3_3', output_layerout[8]) + talker('joint_5_1', output_layerout[12]) + talker('joint_5_2', output_layerout[13]) + talker('joint_5_3', output_layerout[14]) + talker('joint_7_1', output_layerout[18]) + talker('joint_7_2', output_layerout[19]) + talker('joint_7_3', output_layerout[20]) + talker('joint_2_1', output_layerout[3]) + talker('joint_2_2', output_layerout[4]) + talker('joint_2_3', output_layerout[5]) + talker('joint_4_1', output_layerout[9]) + talker('joint_4_2', output_layerout[10]) + talker('joint_4_3', output_layerout[11]) + talker('joint_6_1', output_layerout[15]) + talker('joint_6_2', output_layerout[16]) + talker('joint_6_3', output_layerout[17]) + talker('joint_8_1', output_layerout[21]) + talker('joint_8_2', output_layerout[22]) + talker('joint_8_3', output_layerout[23]) #calculating carolis term to find classifier for KT carolis_term_leg1 = Leg_attribute.carolis_term(leg1.m_l, 9.8, leg1.r_1, leg1.l_1, joint_angle_leg_1) -- GitLab