Method and system for continuous action semantic encoding and translation of humanoid entity robot

Abstract

The application discloses a continuous action semantic coding and translation method and system for a humanoid robot, and the method comprises the following steps: establishing a unified reference time axis, and setting a sampling and coding element action data structure and a decoding element action data structure; sampling data of each actuator according to the sampling and coding element action data structure; coding the sampling points according to action semantics, and outputting the coding and storing the coding in an action library after the coding is completed; taking out the action coding stream from the action library when the robot moves, and decoding the action coding stream; calculating the continuous action coding stream by using an action smoothing algorithm after the decoding, and filling each action point of the continuous action coding stream according to the decoding element action data structure; and outputting the decoded action coding stream and executing the decoded action coding stream according to the unified time axis. The joints of the humanoid robot can independently move at the same time point, and the linkage action of the humanoid robot can be realized on the whole.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of robots, in particular to a continuous action semantic coding and translation method and system of a humanoid entity robot. BACKGROUND

[0002] The robot designed and manufactured by imitating the form and behavior of human beings is a humanoid robot, which generally has human-like limbs and head respectively or simultaneously. The robot is generally designed into different shapes according to different application requirements, such as mechanical arm for industrial use, wheelchair robot, walking robot, etc. The research of humanoid robot integrates multiple sciences such as machinery, electronics, computer, material, sensor and control technology, which represents the development level of high technology of a country. From the research status of robot technology and artificial intelligence, there is still a long way to go to completely realize the humanoid robot with high intelligence and high flexibility, and human beings have not thoroughly understood themselves, which limits the development of humanoid robots.

[0003] The humanoid robot has the appearance of human beings, can adapt to the living and working environment of human beings, replace human beings to complete various tasks, and can expand the ability of human beings in many aspects, and is widely used in many fields such as service, medical treatment, education and entertainment.

[0004] The task of the control system is to control the actuator of the robot according to the operation instruction program of the robot and the signal fed back from the sensor, so that the robot can complete the specified movement and function. If the robot does not have the information feedback feature, the control system is called open-loop control system; if the robot has the information feedback feature, the control system is called closed-loop control system. This part is mainly composed of computer hardware and control software. The software is mainly composed of human-machine interaction system for contacting human beings and robots and control algorithm, etc. The role of this part is equivalent to the brain of human beings.

[0005] Raspberry Pi is a microcomputer designed for computer programming education, which is only the size of a credit card. It is small in size, but has all the basic functions of a PC. If a control system for a robot is designed based on Raspberry Pi, the robot can also be small, light and flexible in action. At present, there is no control system for a humanoid entity robot based on Raspberry Pi. Therefore, it is urgent to provide a control system for a humanoid entity robot based on Raspberry Pi. SUMMARY

[0006] The purpose of the present application is to overcome the defects of the prior art, provide a continuous action semantic coding and translation system and method of a humanoid entity robot, and solve the problem that there is no action semantic coding and translation system of a humanoid entity robot at present to make up for the technical blank.

[0007] The technical solutions adopted by the present application are as follows:

[0008] A continuous action semantic coding and translation method of a humanoid entity robot, comprising:

[0009] establishing a unified reference time axis, and setting a sampling and coding element action data structure and a decoding element action data structure;

[0010] sampling data of each actuator according to the sampling and coding element action data structure;

[0011] coding the sampling points according to action semantics, and outputting the coding and storing it in an action library after coding is completed;

[0012] taking out an action coding stream from the action library when the robot is moving, and decoding the action coding stream;

[0013] calculating a continuous action coding stream by using an action smoothing algorithm after decoding, and filling each action point of the continuous action coding stream according to the decoding element action data structure;

[0014] outputting the decoded action coding stream and executing the decoded action coding stream according to the unified time axis.

[0015] In some embodiments of the method of the present application, the sampling and coding element action data structure is represented as S=[id, t, r]; wherein S represents the sampling and coding data structure, id represents the actuator number, t represents the time point at which the actuator runs to the current angle, and r represents the current angle of the actuator.

[0016] In some embodiments of the method of the present application, the decoding element action data structure is represented as A=[id, t, r, v, a]; wherein A represents the decoding element action, id represents the actuator number, t represents the time point at which the actuator runs to the current angle, r represents the current angle of the actuator, v represents the current speed of the actuator, and a represents the current acceleration of the actuator.

[0017] In some embodiments of the method of the present application, the method further comprises setting a joint element action data structure B=[t, An] during robot movement; wherein B represents the joint action, t represents the unified time axis of the joint action, and An represents the current information of the N actuators.

[0018] In some embodiments of the method of the present application, coding the sampling points according to action semantics comprises:

[0019] setting a unified reference time axis;

[0020] dividing the joint action into N interval segment actions;

[0021] setting the actuator time of the longest time action point as the end time of the joint action;

[0022] Each executor fills data at time point tn for each interval segment action according to meta-action A, n represents the nth interval segment on time axis t, n is a natural number from 1 to N;

[0023] Fill information of all executors at time point t n according to joint action B;

[0024] Determine the end time point of joint action: if the encoding action ends, output the action encoding stream; otherwise, return to continue filling the meta-action A until the encoding action ends.

[0025] In some embodiments of the method, decoding the action encoding stream comprises:

[0026] Setting a unified reference time axis;

[0027] Reading all joint action encoding streams from the action library;

[0028] Taking the action encoding stream at time point t n ;

[0029] Calculating complete continuous smooth action points according to the action encoding stream using an action smoothing algorithm;

[0030] Executing the action encoding stream at time point t n ;

[0031] Determine the end time point of joint action: if the decoding action ends, stop the execution executor; if the decoding does not end, return to continue taking the next time point action encoding stream until the decoding action ends.

[0032] In some embodiments of the method, the action smoothing algorithm uses a spline interpolation algorithm.

[0033] A continuous action semantic encoding and translation system for a humanoid entity robot, which, when executed by a processor, implements the continuous action semantic encoding and translation method for a humanoid entity robot as described above.

[0034] In some embodiments of the system, the system comprises:

[0035] A construction module for establishing a unified reference time axis, and setting a sampling and encoding meta-action data structure and a decoding meta-action data structure;

[0036] A sampling module for sampling data of each executor according to the sampling and encoding meta-action data structure, encoding the sampling points according to the action semantics, and outputting the encoding and storing it in the action library after the encoding is completed;

[0037] A decoding module is configured to take out the action code stream from the action library and decode the action code stream when the robot moves.

[0038] An executing module is configured to calculate the continuous action code stream by using the action smoothing algorithm after decoding, fill each action point of the continuous action code stream according to the decoded action data structure, output the decoded action code stream, and execute the decoded action code stream according to the unified time axis.

[0039] The above technical solution makes the application achieve the following technical effects:

[0040] Under the unified time axis t, the actuators numbered 1 to N rotate according to the respective action code streams, the executing streams of the actuators are independent and do not interfere with each other, the joints of the humanoid robot are independently moved at the same time point, and the linkage action of the humanoid robot is ultimately realized. BRIEF DESCRIPTION OF DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0042] Figure 1 The framework diagram of the continuous action semantic coding and translation system of the humanoid entity robot provided by the embodiment of the present application.

[0043] Figure 2 The executing stream diagram of each actuator under the unified reference time axis provided by the embodiment of the present application.

[0044] Figure 3 The action coding flowchart provided by the embodiment of the present application.

[0045] Figure 4 The action decoding flowchart provided by the embodiment of the present application. DETAILED DESCRIPTION

[0046] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0047] Reference Figure 1The embodiment of the present application provides a continuous action semantic coding and translation system of a humanoid entity robot, and the whole system comprises the following four parts.

[0048] The construction module 11 is used for establishing a unified reference time axis, and setting a sampling and coding element action data structure and a decoding element action data structure.

[0049] The sampling module 12 is used for sampling data of each actuator according to the sampling and coding element action data structure, coding the sampling points according to the action semantics, and outputting the coding and storing in the action library after coding.

[0050] The decoding module 13 is used for taking out the action coding stream from the action library when the robot moves, and decoding the action coding stream.

[0051] The execution module 14 is used for calculating the continuous action coding stream by using the action smoothing algorithm after decoding, filling each action point of the continuous action coding stream according to the decoding element action data structure, and outputting the decoded action coding stream and executing the decoded action coding stream according to the unified time axis.

[0052] The above system can be stored in an electronic device and a computer readable storage medium, and when the system is executed by a processor, a continuous action semantic coding and translation method of a humanoid entity robot is executed, and the specific method steps are as follows.

[0053] Firstly, a unified reference time axis is established, and a sampling and coding element action data structure and a decoding element action data structure are set, so as to have unified data structure standards and time standards when sampling and action coding, decoding and executing the action code stream.

[0054] Then, data of each actuator (a motor or a rudder) is sampled according to the sampling and coding element action data structure.

[0055] Then, the sampling points are coded according to the action semantics, and the coding is output and stored in the action library after coding, and the action coding stream needs to be taken out from the action library when the robot moves, and the action coding stream is decoded.

[0056] After decoding, the continuous action coding stream is calculated by using the action smoothing algorithm (such as a spline interpolation algorithm), and each action point of the continuous action coding stream is filled according to the decoding element action data structure.

[0057] Finally, the decoded action coding stream is output and executed according to the unified time axis. Specifically, the sampling and coding element action data structure is represented as:

[0058] S=[id,t,r]

[0059] Wherein, S represents a sampling and coding data structure, id represents an actuator number, t represents a time point of actuator running to a current angle, and r represents an actuator current angle.

[0060] The decoding element action data structure is represented as:

[0061] A=[id,t,r,v,a]

[0062] Wherein, A represents a decoding element action, id represents an actuator number, t represents a time point of actuator running to a current angle, r represents an actuator current angle, v represents an actuator current rotating speed, and a represents an actuator current acceleration.

[0063] The system is also provided with a joint element action data structure when the robot moves:

[0064] B=[t,An]

[0065] Wherein, B represents a joint action, t represents a unified time axis of the joint action, and An represents current information of N actuators.

[0066] Cooperation Figure 2 As shown in the figure, under the unified time axis t, the actuators numbered 1 to N rotate according to respective action coding streams, and it can be known from the figure that the execution streams of the actuators are independent of each other and do not interfere with each other, so that independent movement of each joint of the humanoid robot at the same time point can be realized, and finally the overall effect can realize the linkage action of the humanoid robot.

[0067] Further, referring to Figure 3 The specific procedure of encoding the sampling points according to the action semantics in the continuous action semantic coding and translation method of the humanoid entity robot provided by the embodiment of the present application is as follows:

[0068] Step 1: Set a unified reference time axis, and all actuator actions are subject to the reference time axis;

[0069] Step 2: Divide the joint action into N interval segment actions;

[0070] Step 3: Set the actuator time of the longest time action point as the end time of the joint action;

[0071] Step 4: Each actuator fills data at the t n time point according to the element action A for each interval segment action, t n represents the n th interval segment on the time axis t, and n is a natural number from 1 to N;

[0072] Step 5: Fill the information of all actuators at the t n (n is a natural number from 1 to N) time point according to the joint action B;

[0073] Step 6: judging the ending time point of the joint action: if the encoding action ends, output the action encoding stream; otherwise, return to step 4 to continue filling the elementary action A until the encoding action ends.

[0074] Further, referring to Figure 4 The specific procedure for decoding the action encoding stream in the continuous action semantic encoding and translation method of the human-simulating entity robot provided by the embodiment of the present application is as follows:

[0075] Step 1: set a unified reference time axis, and all actuator actions are subject to the reference time axis;

[0076] Step 2: read all joint action encoding streams from the action library;

[0077] Step 3: take out the action encoding stream at time point t n t n represents the nth interval segment on the time axis t, and n is a natural number from 1 to N;

[0078] Step 4: calculate the complete continuous smooth action point according to the action encoding stream by using an action smoothing algorithm (such as an interpolation algorithm);

[0079] Step 5: execute the action encoding stream at time point t n (n is a natural number from 1 to N);

[0080] Step 6: judging the ending time point of the joint action: if the decoding action ends, stop executing the actuator; if the decoding has not ended, return to continue taking the next action encoding stream at time point t

[0081] The parts not involved in the present application are the same as or can be implemented by using the prior art. Although the embodiments of the present application have been shown and described, it can be understood by those of ordinary skill in the art that various changes, modifications, replacements and variations can be made to the embodiments without departing from the principles and spirits of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A method for continuous action semantic encoding and translation of a humanoids robot, characterized by, The method comprises the following steps: establishing a unified reference time axis, and setting a sampling and coding element action data structure and a decoding element action data structure; sampling data of each actuator according to the sampling and coding element action data structure; encoding the sampling points according to action semantics, and outputting the encoding and storing it in an action library after the encoding is completed; when the robot is moving, first taking the action encoding stream from the action library, and decoding the action encoding stream; after decoding, calculating a continuous action encoding stream by using an action smoothing algorithm, and filling each action point of the continuous action encoding stream according to the decoding element action data structure; outputting the decoded action encoding stream and executing the decoded action encoding stream according to the unified time axis; the sampling and coding element action data structure is represented as S=[id, t, r]; wherein S represents the sampling and coding data structure, id represents the actuator number, t represents the time point at which the actuator runs to the current angle, and r represents the current angle of the actuator; the decoding element action data structure is represented as A=[id, t, r, v, a]; wherein A represents the decoding element action, id represents the actuator number, t represents the time point at which the actuator runs to the current angle, r represents the current angle of the actuator, v represents the current rotating speed of the actuator, and a represents the current acceleration of the actuator; further comprising setting a joint element action data structure when the robot is moving: B=[t, An]; wherein B represents the joint action, t represents the unified time axis of the joint action, and An represents the current information of the N actuators; encoding the sampling points according to the action semantics, which comprises the following steps: setting a unified reference time axis; dividing the joint action into N interval segments; Each executor fills data at time point t n according to the meta-action A on each interval segment action. n n represents the nth interval segment on the time axis t, and n is a natural number from 1 to N. Fill all actuators with information at time point t n according to joint action B; setting the actuator time of the longest time action point as the end time of the joint action; 2.The continuous action semantic encoding and translation method of the human-simulating entity robot according to claim 1, wherein, judging the end time point of the joint action: if the encoding action is completed, outputting the action encoding stream; otherwise, returning to continue filling the element action A until the encoding action is completed. decoding the action encoding stream, which comprises the following steps: setting a unified reference time axis; Take out t n Time point action encoding stream; reading all joint action encoding streams from the action library; Perform t n Time point action encoding stream; calculating complete continuous smooth action points by using the action smoothing algorithm according to the action encoding stream; 3.The continuous action semantic encoding and translation method of the human-simulating entity robot according to claim 1, wherein, judging the end time point of the joint action: if the decoding action is completed, stopping the execution of the actuator; if the decoding is not completed, returning to continue taking the next time point action encoding stream until the decoding action is completed.

4. A continuous action semantic encoding and translation system for a humanoid entity robot, comprising: The action smoothing algorithm adopts a spline interpolation algorithm.

5. The continuous action semantic encoding and translation system of the humanoid entity robot of claim 4, wherein, The system is executed by a processor to realize the continuous action semantic encoding and translation method of the humanoid entity robot according to any one of claims 1 to 3. The system comprises: a construction module, which is used for establishing a unified reference time axis, and setting a sampling and coding element action data structure and a decoding element action data structure; a sampling module, which is used for sampling data of each actuator according to the sampling and coding element action data structure, and encoding the sampling points according to action semantics, and outputting the encoding and storing it in an action library after the encoding is completed; a decoding module, which is used for first taking the action encoding stream from the action library when the robot is moving, and decoding the action encoding stream; and a smoothing module, which is used for calculating a continuous action encoding stream by using an action smoothing algorithm after decoding, and filling each action point of the continuous action encoding stream according to the decoding element action data structure. The execution module is used for decoding, calculating the continuous action encoding stream by using the action smoothing algorithm, filling each action point of the continuous action encoding stream according to the decoded element action data structure, outputting the decoded action encoding stream and executing the decoded action encoding stream according to the unified time axis.

Images