Foot-ground contact force estimation and terrain classification method of ostrich-imitating robot

Abstract

The invention discloses a foot-ground contact force estimation and terrain classification method of an ostrich-imitating robot. The method comprises the steps that a foot-ground contact force mapping matrix of the ostrich-imitating robot is constructed, force-position hybrid control is carried out, foot-ground contact force is obtained according to the matrix and input torque of a plate spring to a passive joint, and terrain classification information is obtained after the ground contact state is judged; establishing a terrain classification fusion model based on a weighted voting mechanism, and training the model according to terrain classification information; and performing terrain classification according to the terrain classification information and the trained model. The foot-ground contact force is estimated through the joint corner information of the robot, no extra sensor needs to be installed, real-time and accurate foot-ground contact force estimation and contact state obtaining are achieved, the problem that the performance of an ostrich-imitating robot system is reduced or fails due to the fact that a force sensor is prone to being damaged is effectively solved, system configuration is simplified, and the system reliability is improved. Foot-ground contact force estimation and accurate terrain classification of the ostrich-imitating robot can be completed without depending on an external sensor.

Description

Technical Field

[0001] The invention relates to a method for estimating foot-ground contact force and a terrain classification method, and relates to the field of perception of foot-type robots, and in particular to a method for estimating foot-ground contact force and a terrain classification method of an ostrich-like robot. Background Art

[0002] As a kind of mobile robot, the ostrich-like robot has similar passability to biological organisms and excellent environmental adaptability, and is expected to replace humans in many workplaces to complete work tasks. However, the ostrich-like robot cannot currently operate stably in complex terrains, and changes in terrain can cause instability in movement and even damage to components. The adaptability of the ostrich-like robot to complex environments depends on the acquisition of foot-ground contact force and terrain information. The foot-ground contact force is the key information for the ostrich-like robot to perceive the environment and perform compliant control, and the terrain information is the key information for the ostrich-like robot to adapt to the environment. The method of obtaining foot-ground contact force based on force / torque sensors for terrain recognition and impedance control and other compliant control can usually achieve good control effects, but the force / torque sensor is easy to damage and expensive, which greatly affects the robustness and economy of the robot system. In addition, the installation of force / torque sensors on the feet will increase the mass and moment of inertia of the feet, affecting the robot's motion performance. Therefore, it is an urgent problem to study the foot-ground contact force estimation and terrain classification methods without using force / torque sensors. Summary of the invention

[0003] In order to solve the problems existing in the background technology, the present invention provides a method for estimating the foot-ground contact force and classifying terrain of an ostrich-like robot.

[0004] The technical solution adopted by the present invention is:

[0005] The method for estimating the foot-ground contact force and classifying the terrain of the ostrich-like robot of the present invention comprises:

[0006] In the first step, the foot-ground contact force mapping matrix is ​​constructed based on the kinematic model of the ostrich-like robot including joint leaf springs.

[0007] In the second step, the ostrich-like robot is subjected to force-position hybrid control. During the control process, the passive joints of the ostrich-like robot are subjected to torque balance analysis according to the foot-ground contact force mapping matrix to obtain the estimated foot-ground contact force. The ground contact state is determined based on the estimated foot-ground contact force, and then the terrain classification information including slope information and current characteristic information is obtained.

[0008] The third step is to establish a terrain classification fusion model based on a weighted voting mechanism, train the terrain classification fusion model through the current feature information in the terrain classification information, and obtain the trained terrain classification fusion model.

[0009] In the fourth step, when the ostrich-like robot jumps on the terrain to be detected, the same operation as in the second step is performed on the ostrich-like robot to obtain the classification information of the terrain to be detected, and the terrain classification is performed according to the classification information of the terrain to be detected and the trained terrain classification fusion model.

[0010] In the first step, the passive joints of the ostrich-like robot all use large-rigidity leaf springs as flexible components for providing input torque, and the active joints are driven by servo motors; the limb structure of the ostrich-like robot is a planar four-bar parallelogram mechanism with two passive joints, and the leaf spring deformation of the passive joint is obtained by reading and solving the joint encoder of the ostrich-like robot, and then the input torque of the leaf spring to the passive joint is obtained by the leaf spring deformation and stiffness coefficient of the leaf spring.

[0011] In the first step, the foot-ground contact force mapping matrix is ​​as follows:

[0012]

[0013] Among them, F c is the estimated foot-ground contact force of the ostrich-like robot; J P is the mapping matrix of the foot-ground contact force to the passive joint; D P is the mapping matrix from the joint encoder reading to the input torque of the leaf spring on the passive joint; Δθ is the difference between the joint encoder reading and the joint zero position; τ s is the input torque of the leaf spring on the passive joint; L 3 and L 4 are the lengths of the tibia and tarsus of the ostrich-like robot; s 123 and 1234 are the first and second angular sines, s 123 = sin(θ 1 +θ 2 +θ 3 ), s 1234 = sin(θ 1 +θ 2 +θ 3 +θ 4 ), θ 1 ,θ 2 ,θ 3 ,θ 4 They are the rotation angles of the hip joint, knee joint, first passive joint and second passive joint of the ostrich-like robot, c 123 and c 1234 are the first and second angle cosines, c 123 =cos(θ 1 +θ 2 +θ 3 ), c 1234 =cos(θ1 +θ 2 +θ 3 +θ 4 ).

[0014] In the second step, the ostrich-like robot is subjected to force-position hybrid control, and the ostrich-like robot is controlled to perform jumping actions on different terrains with preset periodic trajectories. The ankle joint servo motor is subjected to position control in the air phase, and force control is performed in the ground contact phase, and the output force is set to a smaller value so that the sole of the foot fits the ground; for each jumping action, during the process of the ostrich-like robot jumping and landing, when the estimated foot-ground contact force exceeds the preset ground contact detection threshold, the current moment is taken as the ground contact moment, the ankle joint rotation angle read by the joint encoder at the ground contact moment is taken as the slope information, and the real-time current of the ankle joint motor is obtained at the same time, the first current extreme point of the ankle joint after the ground contact moment is taken as the ankle joint current peak value, and the mean and variance of each ankle joint current peak value under each terrain are taken as the current characteristic information.

[0015] In the third step, the terrain classification fusion model includes a support vector machine SVM (Support Vector Machine) and a back propagation BP (Back Propagation) neural network optimized based on a particle swarm optimization algorithm PSO (Particle Swarm Optimization). The support vector machine SVM and the back propagation BP neural network optimized based on the particle swarm optimization algorithm PSO are integrated using a weighted voting mechanism based on confidence rate. The mapping relationship between current feature information and classified terrain is constructed through the support vector machine SVM and the back propagation BP neural network optimized based on the particle swarm optimization algorithm PSO, thereby establishing a terrain classification fusion model; wherein the back propagation BP neural network optimized based on the particle swarm optimization algorithm PSO uses the weights and thresholds of the neural network as independent variables, and the mean square error of the network as a fitness function, and uses the particle swarm optimization algorithm PSO to optimize the weights and thresholds of the network to train the model.

[0016] The weighted voting mechanism based on confidence rate integrates the support vector machine SVM and the back propagation BP neural network optimized by the particle swarm optimization algorithm PSO. The confidence Y of the integrated terrain classification fusion model is as follows:

[0017]

[0018] in, and The confidence of the support vector machine SVM model and the back propagation BP neural network optimized by the particle swarm optimization algorithm PSO, and It is the confidence of the output results of the support vector machine SVM model and the back-propagation BP neural network optimized by the particle swarm optimization algorithm PSO.

[0019] In the fourth step, the slope is identified by the slope information in the terrain classification information to be detected, and the current characteristic information in the terrain classification information to be detected is input into the trained terrain classification fusion model to identify the category of the current terrain and complete the terrain classification. In specific implementation, other terrains can be sandy and rocky terrain, flat land, soft terrain, etc.

[0020] The terrain classification system of the present invention comprises:

[0021] The data generation layer performs force-position hybrid control on the ostrich-like robot through a given trajectory, so that the ostrich-like robot generates data.

[0022] The data acquisition layer obtains slope information through the joint encoder readings of the ostrich-like robot when the ostrich-like robot performs force-position hybrid control. At the same time, the current data of the joint driver of the ostrich-like robot is collected, and the current feature information is obtained after feature extraction based on the foot-ground contact force estimated by the foot-ground contact force mapping matrix.

[0023] At the data model layer, the terrain classification model is trained according to the current characteristic information, and then the trained terrain classification fusion model is obtained through a weighted voting mechanism based on confidence rate.

[0024] The data analysis layer estimates the slope based on the slope information and identifies other terrains based on the current characteristic information and the trained terrain classification fusion model.

[0025] The electronic device of the present invention is characterized in that it comprises: a memory and a processor coupled to each other, wherein the memory stores program data, and the processor calls the program data to execute the method as described above.

[0026] The computer-readable storage medium of the present invention stores program data thereon, and is characterized in that the program data implements the method described above when executed by a processor.

[0027] The present invention firstly establishes a mapping matrix of foot-ground contact force to passive joints, maps the foot-ground contact force to passive joints, and then solves the passive joint input torque through the joint encoder reading in the parallelogram structure, and obtains the robot foot-ground contact force according to the torque balance equation of the two passive joints based on the passive joint input torque and the foot-ground contact force mapping torque to the passive joint. The foot-ground contact force is obtained based on the aforementioned foot-ground contact force estimation method for ground contact detection, and force-position hybrid control is performed based on the foot-ground contact state, so that the robot completes the set jumping gait on the terrain to be tested and obtains the joint current peak value; data preprocessing is performed to obtain slope and current characteristic information; the slope is identified through the slope information; the mapping relationship between current characteristic information and terrain is constructed through the support vector machine SVM and the back propagation BP neural network optimized by the particle swarm optimization algorithm PSO, and the identification of sandy and rocky land and flat land is completed, and the two terrain classification models are integrated by using the weighted voting mechanism based on confidence to obtain a terrain classification fusion model, which further improves the classification accuracy. The classification of flat land, sloping land, and sandy and rocky land is completed based on the robot's body information.

[0028] The beneficial effects of the present invention are:

[0029] The present invention estimates the foot-ground contact force by utilizing the robot's joint angle information, without the need to install additional sensors, and achieves real-time and accurate estimation of the foot-ground contact force and contact state acquisition, effectively solving the problem of performance degradation or failure of the ostrich-like robot system caused by the fragility of the force sensor. This simplifies the system configuration, reduces hardware costs, and improves the overall stability of the robot system. The present invention can complete the classification of flat land, sloping land, and sandy and gravelly land without the need to install additional sensors. It has been verified that the accuracy of terrain classification is above 90%. An accurate and reliable terrain perception method is provided, which helps to enhance the flexibility and robustness of the ostrich-like robot's movement. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a diagram of a single leg of the ostrich-like robot of the present invention, wherein: Figure 1 (a) is a structural diagram of a single leg of the ostrich-like robot of the present invention, Figure 1 (b) is a schematic diagram of the single-limb structure of the ostrich-like robot of the present invention;

[0031] Figure 2 is a diagram of a foot-ground contact force estimation framework of the present invention;

[0032] Figure 3 It is a framework diagram of the terrain classification system of the present invention;

[0033] Figure 4 is a flow chart of the jumping gait control of the present invention;

[0034] Figure 5 is a flow chart of current information preprocessing of the present invention;

[0035] Figure 6 is a diagram showing the results of the foot-ground contact force estimation experiment of the present invention, Figure 6 (a) is a diagram showing the experimental results of the collision force estimation on flat ground of the present invention. Figure 6 (b) is a graph showing the experimental results of the collision force estimation on a slope of the present invention. Figure 6 (c) is a graph showing the experimental results of the present invention for estimating the impact force on a sandy ground. DETAILED DESCRIPTION

[0036] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein.

[0037] like Figure 1 As shown in (a), the hip joints, knee joints, tibia joints, ankle joints, connecting rods, and leaf springs of the ostrich-like robot's limbs together form a rigid-flexible coupled closed connecting rod mechanism. To facilitate kinematic analysis, the model is simplified as follows: Figure 1 (b) shows a four-bar linkage structure, where points M and C are the centers of the hip joint, N is the center of the knee motor, P is the center of the first passive joint, and a leaf spring s is installed. 1 , A is the center of the second passive joint, and the installed leaf spring BQ is denoted by s 2 , F is the rotation center of the ankle joint. The line MF is perpendicular to the base, and the angle between it and MN is θ 1 , the angle between the extended line of MN and NP is denoted as θ 2 , the angle between the NP extension line and PA is denoted as θ 3 , the angle between the extended line of PA and AF is denoted as θ 4 MN length is recorded as L 1 , NP length is recorded as L 2 , PA length is recorded as L 3 , AF length is recorded as L 4 The origin of the base coordinate coincides with point M, X 0 The axis coincides with the line connecting MF and points to F, Z 0 The axis is perpendicular to the connecting rod plane and points outward, Y 0 The coordinate axis positions of the axes conform to the right-hand system. The coordinates of each joint under the base coordinate are expressed as follows:

[0038]

[0039] For ease of description, the above formula has the following simplified formula: cos(θ i )=c i , cos(θ i +θ j )=c ij ,sin(θ i )=s 1 ,sin(θ i +θ j )=s ij , and so on.

[0040] like Figure 2 As shown, on this basis, the present invention performs the estimation of the foot-ground contact force and the terrain classification of the ostrich-like robot as follows:

[0041] Firstly, based on the kinematic model of the ostrich-like robot including joint leaf springs, the foot-ground contact force mapping matrix is ​​constructed; the passive joints of the ostrich-like robot all use large-stiffness leaf springs as flexible components to provide input torque, and the active joints are driven by servo motors; the limb structure of the ostrich-like robot is a planar four-bar parallelogram mechanism with two passive joints. The leaf spring deformation of the passive joint is obtained by solving the joint encoder readings of the ostrich-like robot, and then the input torque of the leaf spring on the passive joint is obtained through the leaf spring deformation and stiffness coefficient of the leaf spring.

[0042] There are three types of input torques in the planar four-bar linkage, namely, the mapping torque of the foot-ground collision force on each joint, the input torque of the servo motor on the active joint, and the input torque of the elastic element leaf spring on the passive joint. Since the passive joint cannot transmit torque, the leaf spring of the first passive joint does not act on the leaf spring of the second passive joint; the leaf spring of the second passive joint acts on the second passive joint, and the direction of its leaf spring connecting rod force points to the center of the hip joint. This force will not generate torque on the first passive joint but is resisted by the mechanical structure itself, so the leaf spring of the second passive joint does not act on the leaf spring of the first passive joint. Therefore, on the passive joint, only the torque mapping torque of the foot-ground contact force to the passive joint and the input torque of the leaf spring of the passive joint need to be considered.

[0043] For the working scenario of the ostrich-like robot, the deformation of the leaf spring during the working process of the ostrich-like robot and the input torque to the passive joint can be approximately linearly related. Therefore, the mapping matrix D from the joint encoder reading to the passive joint input torque can be obtained. P , the weighted least square method is used for system identification to obtain the leaf spring stiffness, and the input torque can be obtained from the stiffness coefficient and the leaf spring deformation.

[0044] The foot-ground contact force mapping matrix is ​​as follows:

[0045]

[0046] Among them, F c is the estimated foot-ground contact force of the ostrich-like robot; J P D is the mapping matrix of the foot-ground contact force to the passive joint; P is the mapping matrix from the joint encoder reading to the input torque of the leaf spring on the passive joint; Δθ is the difference between the joint encoder reading and the joint zero position; τ s is the input torque of the leaf spring on the passive joint; L 3 and L 4 are the lengths of the tibia and tarsus of the ostrich-like robot; s 123 and 1234 are the first and second angular sines, s 123 = sin(θ 1 +θ 2 +θ 3 ), s 1234 = sin(θ 1 +θ 2 +θ 3 +θ 4 ), θ 1 ,θ 2 ,θ 3 ,θ 4 They are the rotation angles of the hip joint, knee joint, first passive joint and second passive joint of the ostrich-like robot, c 123 and c 1234 are the first and second angle cosines, c 123 =cos(θ 1 +θ 2 +θ 3 ), c 1234 =cos(θ 1 +θ 2 +θ 3 +θ 4 ).

[0047] Based on the kinematic model of the ostrich-like robot, the mapping matrix J of the ostrich-like robot's foot-ground contact force to the passive joint is constructed. P , considering the construction of the mapping matrix of the foot-ground contact force to the passive joint of the ostrich-like robot geometry:

[0048] First, the foot end position can be obtained as follows:

[0049]

[0050] in, is the transformation matrix from the passive joint to the end joint, are the transformation matrices from the passive joint P to the foot end coordinate system N to the passive joint P; xF and F are the horizontal and vertical coordinates of the foot end in the passive joint coordinate system, x n and n are the horizontal and vertical coordinates of the foot end in the terminal coordinate system respectively.

[0051] The mapping matrix of the foot-ground contact force of the ostrich-like robot to the passive joint is obtained by differentiating the passive joint transformation matrix. Based on the kinematics of the ostrich-like robot, the input torque of the passive joint is solved by reading the single-turn absolute encoder of the joint in the parallelogram structure. A torque balance analysis is performed on the passive joint. Considering the mapping torque of the foot-ground collision force on each joint and the input torque of the elastic element leaf spring on the passive joint, the foot-ground contact force is mapped to the passive joint through the mapping matrix, and a torque balance analysis is performed with the input torque of the passive joint to obtain the torque balance equation. The input torque τ of the leaf spring on the passive joint s The mapping matrix from the passive joint encoder readings of the ostrich-like robot to the passive joint input torque can be obtained, so the foot-ground contact force can be obtained through the robot body sensor information, that is, the foot-ground contact force mapping matrix.

[0052] The mapping matrix from joint encoder readings to passive joint input torque is as follows:

[0053]

[0054] Among them, K P and K A are the leaf spring stiffness at the passive joint point P and point A respectively.

[0055] The passive joint input torque is as follows:

[0056]

[0057] Among them, K 1 and K 2 are the stiffness of leaf spring 1 and leaf spring 2 respectively; Δθ 2 , Δθ 3 and Δθ 4 They are the encoder readings of joints 2, 3, and 4 compared to the zero position.

[0058] Then, the ostrich-like robot is subjected to force-position hybrid control. During the control process, the passive joints of the ostrich-like robot are subjected to torque balance analysis according to the foot-ground contact force mapping matrix to obtain the estimated foot-ground contact force, and the ground contact state is judged according to the estimated foot-ground contact force, thereby obtaining the terrain classification information including the slope information and the current characteristic information. The ostrich-like robot is subjected to force-position hybrid control to control the ostrich-like robot to perform jumping actions on different terrains with a preset periodic trajectory, and the ankle joint servo motor is subjected to position control in the air phase, and force control is performed in the ground contact phase, and the output force is set to a smaller value to make the sole of the foot fit the ground. For each jumping action, when the estimated foot-ground contact force exceeds the preset ground contact detection threshold during the jumping and landing process of the ostrich-like robot, the current moment is taken as the ground contact moment. According to the foot-ground contact force range of the robot in the air and at the ground contact moment during the experiment, 1000N is taken as the ground contact threshold, and when the foot-ground contact force is greater than the ground contact threshold, it is regarded as touching the ground. The ankle joint angle read by the joint encoder at the moment of contact is used as the slope information, and the real-time current of the ankle joint motor is obtained at the same time. The first current extreme point of the ankle joint after the moment of contact is used as the ankle joint current peak value, and the mean and variance of each ankle joint current peak value under each terrain are used as the current characteristic information. Figure 4 As shown, when the foot-ground contact force F c Greater than the set threshold F 1 , the robot is in the ground contact phase, the ankle joint motor uses the force control mode and sets the output force to a smaller value, so that the sole of the foot fits the ground; when the foot-ground contact force F c Less than the set threshold F 2 , it is in the air phase, and the ankle joint motor uses the position control mode in the air phase. At the same time, the slope and current characteristic information are collected. Figure 5 The collision time is recorded as the index value, and the index value is used to record the first current extreme point after the collision time from the ankle motor current as the ankle current peak value of each collision, and the mean and variance of the ankle current peak value are used as current characteristic information.

[0059] According to the contact state of the foot and the ground, force-position hybrid control is performed to make the robot complete the set periodic jumping gait on the terrain to be identified, while collecting ankle joint encoder and joint current information. According to the contact state of the foot and the ground, the two main body information of joint motor current and encoder reading are preprocessed to obtain slope information and current characteristic information.

[0060] Then, a terrain classification fusion model based on a weighted voting mechanism is established. The terrain classification fusion model is trained through the current characteristic information in the terrain classification information to obtain a trained terrain classification fusion model. After multiple experiments, the current characteristic information is labeled according to the terrain category to form a training data set: sandy terrain is 1, and non-sandy terrain is 0. The training data set is input into the SVM model and the BP neural network improved based on PSO for training. The BP neural network improved based on PSO is a three-layer 8×8×1 fully connected BP neural network with 2 inputs and 1 output. The weights and thresholds of the BP neural network are optimized by the PSO algorithm. According to the terrain classification accuracy obtained in the experiment, the confidence rate of the terrain classification model is calculated, and a terrain classification fusion model integrating the SVM terrain classification model and the PSO-optimized BP neural network terrain classification model is obtained using a weighted voting mechanism based on the confidence rate. The terrain classification fusion model constructs the mapping relationship between current feature information and classified terrain through support vector machine SVM and back propagation BP neural network optimized by particle swarm optimization algorithm PSO; the back propagation BP neural network optimized by particle swarm optimization algorithm PSO takes the weight and threshold of the neural network as independent variables, and the mean square error of the network as the fitness function, and optimizes the weight and threshold of the network based on particle swarm optimization algorithm PSO to train the model. The confidence Y of the integrated terrain classification fusion model is as follows:

[0061]

[0062] in, and The confidence of the support vector machine SVM model and the back propagation BP neural network optimized by the particle swarm optimization algorithm PSO, and It is the confidence of the output results of the support vector machine SVM model and the back-propagation BP neural network optimized by the particle swarm optimization algorithm PSO.

[0063] Finally, when the ostrich-like robot jumps on the terrain to be detected, the same operation as in the second step is performed on the ostrich-like robot to obtain the terrain classification information to be detected, and the slope information in the terrain classification information to be detected is used to identify the slope, and the current characteristic information in the terrain classification information to be detected is input into the trained terrain classification fusion model to identify the category of the current terrain and complete the terrain classification. In specific implementation, other terrains can be sandy and gravel terrain, flat land, and soft terrain. The current characteristic information is input into the terrain classification fusion model to complete the classification of sandy and gravel terrain: when the model output is greater than or equal to 0.5, the output result is sandy and gravel terrain, otherwise it is non-sandy and gravel terrain. So far, the ostrich-like robot has completed the classification of three typical terrains: flat land, sloped land, and sandy and gravel land based on the ontology information.

[0064] like Figure 3 As shown, the present invention also designs a terrain classification system, including a data generation layer, a data acquisition layer, a data model layer and a data analysis layer. The data generation layer performs force-position hybrid control on the ostrich-like robot through a given trajectory, so that the ostrich-like robot generates data; the given trajectory obtains the joint trajectory through inverse kinematics for the motion control of the ostrich-like robot, and the force control and position control modes are switched when touching the ground to reduce the impact of the foot-ground collision on the robot joint. When the ostrich-like robot performs force-position hybrid control, the data acquisition layer obtains the slope information through the joint encoder reading of the ostrich-like robot, and collects the current data of the joint driver of the ostrich-like robot at the same time, and obtains the current feature information after feature extraction based on the foot-ground contact force estimated by the foot-ground contact force mapping matrix; the data model layer trains the terrain classification model according to the current feature information, and then obtains the trained terrain classification fusion model through a weighted voting mechanism based on the confidence rate; the data analysis layer estimates the slope according to the slope information, and identifies other terrains according to the current feature information and the trained terrain classification fusion model.

[0065] The present invention first establishes a mapping matrix of the foot-ground contact force to the passive joint, maps the foot-ground contact force to the torque of the passive joint, calculates the deformation of the passive joint flexible component through the encoder reading, and then obtains the passive joint input torque, and performs torque balance analysis on the input torque at the passive joint to obtain the foot-ground contact force. The ground contact detection is performed based on the obtained foot-ground contact force. The robot is subjected to force-position hybrid control based on the foot-ground contact state, so that it completes the set jumping gait on the terrain to be tested, and the joint current and encoder reading of the robot are obtained. Flat land and sloping land are classified according to the ankle joint encoder reading at the moment of contact, and the mean and variance of the ankle joint current peak are obtained as current feature information based on the foot-ground contact state, and the classification of sloping land and sandy and gravelly land is completed through the terrain classification fusion model. The present invention can complete the foot-ground contact force estimation and terrain classification of the ostrich-like robot without relying on external sensors.

[0066] In a specific embodiment of the present invention, a limb prototype platform was used to conduct foot-ground contact force estimation and terrain recognition experiments on three types of terrain: flat land, sloping land, and sandy land. The experiments are as follows:

[0067] 1) Control the robot to complete the set jumping gait on three types of terrain, and record the terrain type and experimental data. The experimental data for contact force estimation are as follows: Figure 6 (a) Figure 6 (b) and Figure 6 As shown in (c), it can be seen from the figure that the estimated contact force value quickly reaches more than 1000 N after touching the ground. Therefore, according to this contact force estimation method, accurate touchdown detection can be performed.

[0068] 2) Conduct multiple experiments under three types of terrain, control the robot to complete the set jumping gait, and record the terrain type and experimental data. Divide the experimental data into training data sets and test data sets, and conduct terrain recognition experiments after training and testing the terrain classification model. The experimental process and results are shown below:

[0069] 1. First, train the terrain classification model. The support vector machine SVM classifier parameters are: support vector machine SVM type: classification support vector machine SVM; kernel function type: linear kernel function; penalty parameter: 1; whether to use heuristics: enabled; termination criterion: 0.001. The current characteristic information is labeled according to the terrain category: sandy terrain is 1, and non-sandy terrain is 0. After training the terrain classification model with a training data set composed of multiple sets of labeled current characteristic information, the identification of sandy and flat land can be completed. The 150 sets of data are divided into training data sets and test data sets at a ratio of 4:1, and the accuracy rates on the two data sets are 100% and 93.3% respectively; for the three-layer 8×8×1 fully connected BP neural network terrain classification model optimized by PSO, the 150 sets of data are divided into training data sets and test data sets at a ratio of 4:1 for training and testing, and the accuracy rates on the two data sets are 100% and 96.7% respectively; the terrain classification fusion model has an accuracy of 100% in the test data set, and the output of the terrain classification fusion model is:

[0070]

[0071] in, and is the confidence of the support vector machine SVM model and the back propagation BP neural network optimized by the particle swarm optimization algorithm PSO. Here, the accuracy of the two models on the experimental test set is used as the model confidence. and It is the confidence of the output results of the support vector machine SVM model and the back-propagation BP neural network optimized by the particle swarm optimization algorithm PSO.

[0072] 2. After the model training is completed, 6 groups of terrain classification experiments are carried out on three types of terrain: slope, flat land and sandy land, and the experimental results are analyzed:

[0073] First, the slope terrain is identified, and the slope terrain identification threshold is set to 15°. The experimental data and results are shown in Table 1. It can be seen that in all six groups of experiments, the slope terrain was correctly identified.

[0074] Table 1 Summary of experimental data on slope identification

[0075]

[0076] Next, we classified the sandy terrain and flat ground, and calculated the mean and variance of the ankle joint motor current peak value during each ground contact according to the foot-ground contact state. The experimental data for flat ground are shown in Table 2:

[0077] Table 2 Summary of flat ground experimental data

[0078]

[0079] The experimental data of sandstone underground are shown in Table 3:

[0080] Table 3 Summary of experimental data on sandy and rocky land

[0081]

[0082] The model input parameters are input into the trained terrain classification model. From the results in Table 2, Table 3 and the test data set, it can be seen that all three models have completed the classification of terrain, and the accuracy of the fusion model result is higher than that of the terrain classification model using support vector machine SVM and BP neural network based on PSO optimization alone.

[0083] In summary, the method of the present invention can accurately estimate the foot-ground contact force and complete the classification of common terrains such as slopes, flat land and sandy land. In addition, the contact force estimation and terrain classification of the method are performed online, which can meet the real-time control requirements of the ostrich-like robot. Then, the contact force estimation and terrain recognition experiments were carried out on the experimental platform. The contact force estimation experiment verified the effectiveness of force estimation and the accuracy of touchdown detection based on force estimation; the terrain classification experiment was able to correctly classify the terrain in 6 experiments conducted under three terrains, verifying the accuracy of the proposed terrain classification method and integrating the two terrain classification models by using a confidence-based weighted voting mechanism to obtain a terrain classification fusion model, which further improved the classification accuracy. This proves that the proposed contact force estimation and terrain classification method of the ostrich-like robot has good accuracy and reliability, and can be applied to the robot's perception system to estimate the foot-ground contact force and classify the terrain.

[0084] The above contents are only the technical ideas of the present invention and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical ideas proposed by the present invention shall fall within the protection scope of the claims of the present invention.

Claims

1. A method for estimating the foot-ground contact force and classifying terrain of an ostrich-like robot, characterized in that: include: The first step is to construct the foot-ground contact force mapping matrix based on the kinematic model of the ostrich-like robot including joint leaf springs; The second step is to perform force-position hybrid control on the ostrich-like robot. During the control process, the passive joints of the ostrich-like robot are subjected to torque balance analysis according to the foot-ground contact force mapping matrix to obtain an estimated foot-ground contact force. The ground contact state is determined according to the estimated foot-ground contact force, and then the terrain classification information including slope information and current characteristic information is obtained. The third step is to establish a terrain classification fusion model based on a weighted voting mechanism, train the terrain classification fusion model through the current feature information in the terrain classification information, and obtain a trained terrain classification fusion model; In the fourth step, when the ostrich-like robot jumps on the terrain to be detected, the same operation as in the second step is performed on the ostrich-like robot to obtain the classification information of the terrain to be detected, and the terrain classification is performed according to the classification information of the terrain to be detected and the trained terrain classification fusion model.

2. The method for estimating the foot-ground contact force and classifying terrain of an ostrich-like robot according to claim 1, characterized in that: include: In the first step, the passive joints of the ostrich-like robot all use leaf springs as flexible components to provide input torque, and the active joints are driven by servo motors; The limb structure of the ostrich-like robot is a planar four-bar parallelogram mechanism with two passive joints. The leaf spring deformation of the passive joint is obtained by solving the readings of the joint encoder of the ostrich-like robot, and then the input torque of the leaf spring on the passive joint is obtained through the leaf spring deformation and stiffness coefficient of the leaf spring.

3. The method for estimating the foot-ground contact force and classifying terrain of an ostrich-like robot according to claim 2, characterized in that: include: In the first step, the foot-ground contact force mapping matrix is ​​as follows: Among them, F c is the estimated foot-ground contact force of the ostrich-like robot; J P D is the mapping matrix of the foot-ground contact force to the passive joint; P is the mapping matrix from the joint encoder reading to the input torque of the leaf spring on the passive joint; Δθ is the difference between the joint encoder reading and the joint zero position; τ s is the input torque of the leaf spring on the passive joint; L3 and L4 are the lengths of the tibia and tarsus of the ostrich-like robot, respectively; s 123 and 1234 are the first and second angular sines, s 123 =sin(θ1+θ2+θ3), s 1234 = sin(θ1+θ2+θ3+θ4), θ1, θ2, θ3, θ4 are the rotation angles of the hip joint, knee joint, first passive joint and second passive joint of the ostrich-like robot, c 123 and c 1234 are the first and second angle cosines, c 123 =cos(θ1+θ2+θ3), c 1234 =cos(θ1+θ2+θ3+θ4).

4. The method for estimating the foot-ground contact force and classifying terrain of an ostrich-like robot according to claim 1, characterized in that: include: In the second step, the ostrich-like robot is subjected to force-position hybrid control, and the ostrich-like robot is controlled to perform jumping actions on different terrains with preset periodic trajectories. For each jumping action, when the estimated foot-ground contact force exceeds the preset touchdown detection threshold during the process of the ostrich-like robot jumping and landing, the current moment is taken as the touchdown moment, and the ankle joint angle read by the joint encoder at the touchdown moment is taken as the slope information. At the same time, the real-time current of the ankle joint motor is obtained, and the first current extreme point of the ankle joint after the touchdown moment is taken as the ankle joint current peak value, and the mean and variance of each ankle joint current peak value under each terrain are taken as the current characteristic information.

5. The method for estimating the foot-ground contact force and classifying terrain of an ostrich-like robot according to claim 1, characterized in that: include: In the third step, the terrain classification fusion model includes a support vector machine SVM and a back propagation BP neural network optimized based on a particle swarm optimization algorithm PSO. The support vector machine SVM and the back propagation BP neural network optimized based on the particle swarm optimization algorithm PSO are integrated using a weighted voting mechanism based on confidence rate. The mapping relationship between current feature information and classified terrain is constructed through the support vector machine SVM and the back propagation BP neural network optimized based on the particle swarm optimization algorithm PSO, thereby establishing a terrain classification fusion model.

6. The method for estimating the foot-ground contact force and classifying terrain of an ostrich-like robot according to claim 5, characterized in that: include: The weighted voting mechanism based on confidence rate integrates the support vector machine SVM and the back propagation BP neural network optimized by the particle swarm optimization algorithm PSO. The confidence Y of the integrated terrain classification fusion model is as follows: in, and The confidence of the support vector machine SVM model and the back propagation BP neural network optimized by the particle swarm optimization algorithm PSO, and It is the confidence of the output results of the support vector machine SVM model and the back-propagation BP neural network optimized by the particle swarm optimization algorithm PSO.

7. The method for estimating the foot-ground contact force and classifying terrain of an ostrich-like robot according to claim 1, characterized in that: include: In the fourth step, the slope is identified by the slope information in the terrain classification information to be detected, and the current characteristic information in the terrain classification information to be detected is input into the trained terrain classification fusion model to identify the category of the current terrain and complete the terrain classification.

8. A terrain classification system suitable for the method according to any one of claims 1 to 7, characterized in that: include: The data generation layer performs force-position hybrid control on the ostrich-like robot through a given trajectory, so that the ostrich-like robot generates data; The data acquisition layer, when the ostrich-like robot performs force-position hybrid control, obtains the slope information through the joint encoder readings of the ostrich-like robot, and simultaneously collects the current data of the joint driver of the ostrich-like robot, and obtains the current feature information after performing feature extraction based on the foot-ground contact force estimated by the foot-ground contact force mapping matrix; In the data model layer, the terrain classification model is trained according to the current characteristic information, and then the trained terrain classification fusion model is obtained through a weighted voting mechanism based on confidence rate; The data analysis layer estimates the slope based on the slope information and identifies other terrains based on the current characteristic information and the trained terrain classification fusion model.

9. An electronic device, characterized in that: include: A memory and a processor coupled to each other, wherein the memory stores program data, and the processor calls the program data to execute the method according to any one of claims 1 to 8.

10. A computer-readable storage medium having program data stored thereon, characterized in that: When the program data is executed by a processor, the method according to any one of claims 1 to 8 is implemented.

Images