control device

Abstract

To provide a control device capable of enhancing versatility.SOLUTION: A control part 10 controls a machine device provided with driving parts Aa and Ab. The control device 10 includes detection parts Da and Db, processing parts Pa, Pb and Pc, and driving control parts Ca and Cb. The detection parts Da and Db convert a detection signal output from sensors Sa and Sb into sensory information, and transmits the converted sensory information to communication networks Na and Nb. The processing parts Pa, Pb and Pc analyze the sensory information transmitted to the communication networks Na and Nb, generate action command information, and transmit the generated action command information to the communication networks Na and Nb. The driving control parts Ca and Cb analyze the action command information transmitted to the communication networks Na and Nb, and control the driving parts Aa and Ab on the basis of the analyzed action command information.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a control device.

Background Art

[0002] Conventionally, there is a robot described in Patent Document 1 below. This robot includes a tactile sensing unit, a generation unit, a storage unit, an analysis unit, and a control unit. The tactile sensing unit detects external contact. The generation unit generates contact information based on the detection result. The storage unit accumulates and stores the contact information for a predetermined period. The analysis unit analyzes a tendency for each user based on the accumulated contact information. The control unit controls a mechanism for causing the robot to perform an action, and dynamically changes the action based on the analysis result.

Prior Art Literature

Patent Literature

[0003]

Patent Document 1

Summary of the Invention

Problem to be Solved by the Invention

[0004] In the control device described in Patent Document 1, there is room for improvement in terms of versatility.

[0005] The present invention has been made in view of such circumstances, and an object of the present invention is to provide a control device capable of improving versatility.

Means for Solving the Problem

[0006] A control device that solves the above problems is a control device for controlling a mechanical device equipped with multiple drive units, and comprises a detection unit, a processing unit, and a drive control unit. The detection unit converts a detection signal output from a detection unit sensor, which is provided on the mechanical device, into sensory information when the detection unit sensor detects predetermined information, and transmits the converted sensory information to a transmission line. The processing unit analyzes the sensory information transmitted from the detection unit to the transmission line, and based on the analysis results of the sensory information, generates action command information, which is information about the action that the mechanical device should take, and transmits the generated action command information to the transmission line. The drive control unit analyzes the action command information transmitted from the processing unit to the transmission line and controls the drive units based on the analyzed action command information.

[0007] With this configuration, the detection unit, processing unit, and drive control unit operate based on sensory information and behavioral command information, making it more versatile compared to conventional control devices that operate with numerical signals or complex control signals. [Effects of the Invention]

[0008] The control device of the present invention makes it possible to increase versatility. [Brief explanation of the drawing]

[0009] [Figure 1] A block diagram showing the schematic configuration of the control device of the first embodiment. [Figure 2] A schematic diagram showing the configuration of the robot device according to the first embodiment. [Figure 3] A block diagram showing part of the configuration of the control device of the first embodiment. [Figure 4] A block diagram showing part of the configuration of the control device of the first embodiment. [Figure 5] A block diagram showing the configuration of the detection unit of the first embodiment. [Figure 6] A block diagram showing the configuration of the neural network in the first embodiment. [Figure 7] A block diagram showing the configuration of the memory controller of the first embodiment. [Figure 8] A diagram schematically illustrating an example of information included in a semantic library according to the first embodiment. [Figure 9] A block diagram showing a configuration of a processing unit according to the first embodiment. [Figure 10] A diagram schematically illustrating an example of information included in a semantic library according to the first embodiment. [Figure 11] A block diagram showing a configuration of a drive control unit according to the first embodiment. [Figure 12] A block diagram showing a configuration of a processing unit according to the second embodiment. [Figure 13] A block diagram showing a configuration of a drive control unit according to the second embodiment. [Figure 14] A block diagram showing a partial configuration of each of a detection unit and a processing unit according to the third embodiment. [Figure 15] A block diagram showing a partial configuration of each of a processing unit and a drive control unit according to the third embodiment. DETAILED DESCRIPTION OF EMBODIMENTS

[0010] Hereinafter, an embodiment of a control device will be described with reference to the drawings. To facilitate understanding of the description, the same reference numerals are given to the same constituent elements in each drawing as much as possible, and duplicate descriptions are omitted.

[0011] <First Embodiment> First, an outline of the control device according to the first embodiment will be described.

[0012] (Outline of Control Device) As shown in FIG. 1, the control device 10 of the present embodiment is mounted on a mechanical device 20 that is a control target thereof. The mechanical device 20 is, for example, a humanoid robot, an industrial robot, or the like. The control device 10 includes a plurality of high-speed detection units Da, a plurality of low-speed detection units Db, a plurality of high-speed drive control units Ca, a plurality of low-speed drive control units Cb, a plurality of upper processing units Pa, a plurality of middle processing units Pb, and a plurality of lower processing units Pc.

[0013] Multiple high-speed detection units Da, multiple high-speed drive control units Ca, multiple upper-level processing units Pa, and multiple intermediate-level processing units Pb are connected to each other via a high-speed communication network Na. The information transmitted to the high-speed communication network Na is broadcast information and is processed arbitrarily by the multiple high-speed detection units Da, multiple high-speed drive control units Ca, multiple upper-level processing units Pa, and multiple intermediate-level processing units Pb.

[0014] Multiple low-speed detection units Db, multiple low-speed drive control units Cb, multiple upper-level processing units Pa, multiple intermediate-level processing units Pb, and multiple lower-level processing units Pc are connected to each other via a low-speed communication network Nb. The information transmitted to the low-speed communication network Nb is also broadcast information and is processed arbitrarily by the multiple low-speed detection units Db, multiple low-speed drive control units Cb, multiple upper-level processing units Pa, multiple intermediate-level processing units Pb, and multiple lower-level processing units Pc.

[0015] The high-speed detection unit Da and the low-speed detection unit Db are parts of the mechanical device 20 that detect predetermined physical quantities. For this purpose, the high-speed detection unit Da and the low-speed detection unit Db each have sensors Sa and Sb, respectively, for detecting predetermined physical quantities. Sensors Sa and Sb are mounted on the mechanical device 20 and include, for example, camera sensors, image sensors, acoustic sensors, posture sensors, taste sensors, radio wave sensors, gas sensors, temperature sensors, contact sensors, and pressure sensors. Sensor Sa mounted on the high-speed detection unit Da is required to operate faster than sensor Sb mounted on the low-speed detection unit Db. Whether various sensors are classified into the high-speed detection unit Da or the low-speed detection unit Db depends on the configuration of the mechanical device 20, which is controlled by the control device 10. For example, if the mechanical device 20 is a humanoid robot, it may be equipped with camera sensors to acquire information about the surroundings and contact sensors to detect whether or not an object has come into contact with the robot. In this case, since the image information captured by the camera sensor is used to control the entire robot, it is desirable that this image information be processed at a higher speed. Therefore, the camera sensor is classified as one of the sensors Sa of the multiple high-speed detection unit Da. On the other hand, contact information detected by the contact sensor is rarely used as high-priority information when controlling the robot. Therefore, the contact sensor is classified as one of the sensors Sb of the multiple low-speed detection unit Db.

[0016] The detection units Da and Db convert detection signals output from sensors Sa and Sb when sensors Sa and Sb detect predetermined physical quantities into sensory information, abstract the converted sensory information, and transmit it to the high-speed communication network Na. For example, if sensor Sb is a contact sensor and is located on the robot's right hand, the low-speed detection unit Db generates verbalized sensory information such as "the robot's right hand is touching a hard object" or "the robot's right hand is touching a cold object" based on the detection signal from sensor Sb. In this embodiment, the process of verbalizing sensory information corresponds to the process of abstracting sensory information. The low-speed detection unit Db then transmits the verbalized sensory information to the low-speed communication network Nb. The sensory information includes information representing the five human senses and information about the location where the sensation was obtained. The high-speed detection unit Da also transmits sensory information corresponding to the detection signal to the high-speed communication network Na when sensor Sa outputs a detection signal after detecting a predetermined physical quantity.

[0017] Sensory information transmitted from the high-speed detection unit Da to the high-speed communication network Na is acquired by multiple higher-level processing units Pa and multiple intermediate-level processing units Pb. Similarly, sensory information transmitted from the low-speed detection unit Db to the low-speed communication network Nb is acquired by multiple higher-level processing units Pa, multiple intermediate-level processing units Pb, and multiple lower-level processing units Pc. Each processing unit Pa, Pb, and Pc analyzes the sensory information acquired via the respective communication networks Na and Nb, and generates action command information, which is information about the actions that the machine device 20 should take, based on the analysis results. The action command information includes information that defines the overall actions of the machine device 20, information that defines the operation of each part of the machine device 20, and so on.

[0018] The upper-level processing unit Pa, the intermediate-level processing unit Pb, and the lower-level processing unit Pc each perform different processing tasks. For example, if the mechanical device 20 controlled by the control device 10 is a humanoid robot, the upper-level processing unit Pa generates action command information that determines the actions the entire robot should take, such as the direction in which the mechanical device 20 should move. The intermediate-level processing unit Pb and the lower-level processing unit Pc generate action command information that determines the actions of multiple parts of the robot. For example, multiple intermediate-level processing units Pb generate action command information that determines the actions of multiple hands of the robot. Also, multiple lower-level processing units Pc generate action command information that determines the actions of multiple arms of the robot. The processing of the intermediate-level processing unit Pb takes precedence over the processing of the lower-level processing unit Pc. Furthermore, the processing of the upper-level processing unit Pa takes precedence over the processing of both the lower-level processing unit Pc and the intermediate-level processing unit Pb. One of the multiple upper-level processing units Pa, multiple intermediate-level processing units Pb, and multiple lower-level processing units Pc can exchange various information with an external device De such as a random number generator.

[0019] For example, if the low-speed detection unit Db transmits sensory information such as "the robot's right hand is touching a cold object" to the low-speed communication network Nb, and this sensory information is received by the intermediate processing unit Pb, the intermediate processing unit Pb analyzes the sensory information and generates action command information such as "immediately move the robot's right hand away from the object." The intermediate processing unit Pb then transmits the generated action command information to the high-speed communication network Na and the low-speed communication network Nb. The other upper-level processing unit Pa and lower-level processing unit Pc operate similarly based on the sensory information transmitted from the detection units Da and Db to the high-speed communication network Na and the low-speed communication network Nb.

[0020] Action command information transmitted from the upper-level processing unit Pa and the intermediate-level processing unit Pb to the high-speed communication network Na is received by multiple high-speed drive control units Ca. Action command information transmitted from the upper-level processing unit Pa, the intermediate-level processing unit Pb, and the lower-level processing unit Pc to the low-speed communication network Nb is received by multiple low-speed drive control units Cb. The high-speed drive control units Ca control the high-speed drive unit Aa provided in the mechanical device 20. The low-speed drive control units Cb control the low-speed drive unit Ab provided in the mechanical device 20. The drive units Aa and Ab are motor devices, speaker devices, and lighting devices, etc., that constitute the mechanical device 20. The high-speed drive unit Aa is a device that requires faster operation than the low-speed drive unit Ab. Whether various devices provided in the mechanical device 20 are classified as high-speed drive unit Aa or low-speed drive unit Ab depends on the configuration of the mechanical device 20 that is controlled by the mechanical device 20. For example, if the mechanical device 20 is a humanoid robot, the actuator device that operates the robot's hand is classified as a high-speed drive unit Aa because high-speed operation is required. On the other hand, the actuator device that operates the robot's arm is classified as a low-speed drive unit Ab because high-speed operation is not required.

[0021] The drive control units Ca and Cb control the drive units Aa and Ab, which are the targets of control, based on the action command information transmitted from the respective processing units Pa, Pb, and Pc to the communication networks Na and Nb. For example, if the intermediate processing unit Pb transmits action command information such as "immediately move the robot's right hand away from the object" to the high-speed communication network Na, and this action command information is received by the high-speed drive control unit Ca, the high-speed drive control unit Ca analyzes the action command information and operates the high-speed drive unit Aa, which is an actuator device that moves the robot's hand.

[0022] (Specific configuration of machinery and control devices) Next, we will describe the specific configuration of the control device 10 when the mechanical device 20 shown in Figure 1 is a humanoid robot device.

[0023] As shown in Figure 2, the robot device 30 of this embodiment comprises a body 31, a head 32, shoulders 33a, 33b, arms 34a, 34b, hands 35a, 35b, waist 36, and legs 37a, 37b. The robot device 30 is powered by a battery 38 built into the waist 36.

[0024] The head 32 is equipped with a camera sensor 40, a distance sensor 41, a sound sensor 42, and a speaker device 43. The camera sensor 40 captures images of the area in front of the robot and acquires image data of the area in front of the robot. The distance sensor 41 is a sensor that measures the distance to objects in front of the robot. The sound sensor 42 is a sensor that detects sounds in the robot's surroundings. The speaker device 43 functions as the robot's mouth and emits sounds, music, etc.

[0025] The shoulder sections 33a and 33b each contain actuator devices 330a and 330b, respectively. The actuator devices 330a and 330b operate the arm sections 34a and 34b, respectively. In this embodiment, the actuator devices 330a and 330b correspond to the drive units.

[0026] Contact sensors 50-53 are provided on the arms 34a and 34b, respectively. Contact sensors 54 and 55 are also provided on the hands 35a and 35b, respectively. Furthermore, contact sensors 56-59 are also provided on the legs 37a and 37b, respectively. Contact sensors 50-59 are capable of detecting whether or not an object has made contact with them. In addition, contact sensors 50-59 can detect the physical properties of the object they have contacted, such as elastic properties and viscous properties, and output a detection signal corresponding to the detected physical properties.

[0027] The hand sections 35a and 35b each have an actuator device 350a and 350b built into them. The actuator devices 350a and 350b operate the hand sections 35a and 35b, respectively.

[0028] Next, with reference to Figures 3 and 4, the specific configuration of the control device 10 mounted on the robot device 30 will be described.

[0029] The control device 10 includes a high-speed communication network Na as shown in Figure 3 and a low-speed communication network Nb as shown in Figure 4.

[0030] As shown in Figure 3, the high-speed communication network Na is connected to the high-speed detection units Da1 to Da5, the high-speed drive control units Ca1 to Ca3, the upper-level processing units Pa1 to Pa3, and the intermediate-level processing units Pb1 and Pb2.

[0031] The high-speed detection units Da1 to Da5 process the detection signals from the camera sensor 40, distance sensor 41, sound sensor 42, and contact sensors 54 and 55, respectively. The high-speed drive control units Ca1 to Ca3 control the speaker device 43 and the actuator devices 350a and 350b, respectively.

[0032] The high-speed detection unit Da1 generates sensory information related to the vision of the robot device 30 based on image information acquired by the camera sensor 40, and transmits the generated sensory information to the high-speed communication network Na. Sensory information related to vision is information such as "the signal in front of the robot's eyes is red."

[0033] The high-speed detection unit Da2 generates sensory information related to the robot device 30's vision based on distance information to a predetermined object acquired by the distance measuring sensor 41, and transmits the generated sensory information to the high-speed communication network Na. Sensory information related to vision is, for example, information such as "There is an obstacle 5 meters ahead of the robot."

[0034] The high-speed detection unit Da3 generates sensory information related to the robot's hearing based on the audio information acquired by the audio sensor 42, and transmits the generated sensory information to the high-speed communication network Na. Sensory information related to hearing is, for example, information such as "I can hear the sound of a train."

[0035] The high-speed detection units Da4 and Da5 generate sensory information related to the tactile sense of the robot device 30 based on contact information acquired by the contact sensors 54 and 55, and transmit the generated sensory information to the high-speed communication network Na. Sensory information related to the tactile sense is, for example, information such as "the robot's right hand is in contact with a hard object."

[0036] The higher-level processing units Pa1 to Pa3 generate action command information based on sensory information transmitted from the high-speed detection units Da1 to Da5 to the high-speed communication network Na, and transmit the generated action command information to the high-speed communication network Na. For example, the higher-level processing unit Pa1 generates action command information that defines the actions of the entire robot device 30 by using the sensory information transmitted from the high-speed detection units Da1 to Da5. The higher-level processing unit Pa2 also generates action command information that defines the actions of the entire robot device 30 by using the sensory information transmitted from the high-speed detection units Da1 to Da5. The higher-level processing unit Pa1 generates action command information that is more related to the actions of the robot device 30 than the higher-level processing unit Pa2. Since the higher-level processing unit Pa1 is positioned higher than the higher-level processing unit Pa2, the action command information of the higher-level processing unit Pa1 takes precedence over the action command information of the higher-level processing unit Pa2. The higher-level processing unit Pa3 generates action command information related to voice, for example, by using the sensory information transmitted from the high-speed detection units Da1 to Da5.

[0037] The right-hand intermediate processing unit Pb1 generates action command information that defines the actions of the right hand portion 35a of the robot device 30 by using sensory information transmitted from the high-speed detection units Da1 to Da5, and transmits the generated action command information to the high-speed communication network Na. The left-hand intermediate processing unit Pb2 generates action command information that defines the actions of the left hand portion 35b of the robot device 30 by using sensory information transmitted from the high-speed detection units Da1 to Da5, and transmits the generated action command information to the high-speed communication network Na.

[0038] The high-speed drive control unit Ca1 controls the speaker device 43 based on the voice-related action command information among the action command information transmitted from each processing unit Pa1 to Pa3, Pb1, and Pb2 to the high-speed communication network Na.

[0039] The high-speed drive control unit Ca2 controls the actuator device 350a of the right hand section 35a based on the action command information related to the right hand, which is transmitted from each processing unit Pa1 to Pa3, Pb1, and Pb2 to the high-speed communication network Na.

[0040] The high-speed drive control unit Ca3 controls the actuator device 350b of the left hand section 35b based on the action command information related to the left hand, which is transmitted from each processing unit Pa1 to Pa3, Pb1, and Pb2 to the high-speed communication network Na.

[0041] As shown in Figure 4, the low-speed communication network Nb is connected to the low-speed detection units Db1 to Db8, the low-speed drive control units Cb1 and Cb2, and the lower-level processing units Pc1 to Pc3.

[0042] The low-speed detection units Db1 to Db8 generate sensory information related to the tactile sense of the robot device 30 based on contact information acquired by contact sensors 50 to 53 and 56 to 59, and transmit the generated sensory information to the high-speed communication network Na. Sensory information related to the tactile sense is, for example, information such as "the robot's right foot is in contact with a soft object."

[0043] The lower-level processing units Pc1 to Pc3 generate action command information that determines the actions of the arms 34a and 34b of the robot device 30 based on sensory information transmitted from contact sensors 50 to 53 and 56 to 59, and transmit the generated action command information to the low-speed communication network Nb.

[0044] The low-speed drive control units Cb1 and Cb2 control the actuator devices 330a and 330b provided on the shoulder units 33a and 33b, respectively, based on the action command information related to the arms 34a and 34b, which is transmitted from each of the lower-level processing units Pc1 to Pc3 to the low-speed communication network Nb.

[0045] (Configuration of the detection unit) Next, the configurations of the detection units Da1 to Da5 and Db1 to Db8 will be described in detail. Since each of the detection units Da1 to Da5 and Db1 to Db8 has the same or similar configuration, the configuration of the high-speed detection unit Da4, which corresponds to the contact sensor 54 of the right hand portion 35a, will be described as representative below.

[0046] As shown in Figure 5, the high-speed detection unit Da4 comprises a pre-processing unit 70, a recognition unit 71, a learning unit 72, an estimation unit 73, a determination unit 74, an assignment unit 75, a language processing unit 76, a data output unit 77, a communication processing unit 78, a data receiving unit 79, a writing processing unit 80, a data compression unit 81, and a memory controller 82. The recognition unit 71, the learning unit 72, the estimation unit 73, the memory controller 82, and the language processing unit 76 are connected to each other via a data-memory bus 83 so as to be able to communicate with each other.

[0047] The preprocessing unit 70 performs preprocessing on the detection signal output from the contact sensor 54. Preprocessing includes filtering, etc.

[0048] The recognition unit 71 recognizes contact information detected by the contact sensor 54 by analyzing the detection signal of the contact sensor 54, which has been pre-processed by the pre-processing unit 70. The recognition unit 71 is composed of a neural network NN as shown in Figure 6. As shown in Figure 6, the neural network NN comprises a plurality of input layers 1000, a plurality of intermediate layers 1001, and a plurality of output layers 1002. The detection signal of the pre-processed contact sensor 54 is input to the input layer 1000. The intermediate layer 1001 acquires information from the input layer 1000 and performs various calculations based on the acquired information. The output layer 1002 outputs a plurality of data processed with an activation function by applying predetermined weights to the input layer 1000 and the intermediate layer 1001. The output layer 1002 of the recognition unit 71 outputs physical property information detected by the contact sensor 54. This detected physical property information includes, for example, numerical data such as hardness, elastic properties, and viscosity.

[0049] The learning unit 72 sequentially learns the correspondence between the detection signal of the contact sensor 54, which has been pre-processed by the pre-processing unit 70, and the physical property information detected by the contact sensor 54. The learning unit 72 acquires the physical property information detected by the contact sensor 54 from the determination unit 74. The learning unit 72 is composed of a neural network NN as shown in Figure 6. The learning unit 72 outputs estimated physical property information by using the learning information as input information for the detection signal of the contact sensor 54, which has been pre-processed by the pre-processing unit 70.

[0050] The estimation unit 73 outputs estimated physical property information by using machine learning or a computer, etc., with the detection signal from the contact sensor 54, which has been pre-processed by the pre-processing unit 70, as input information.

[0051] The determination unit 74 calculates the final detected physical property information by combining the detected physical property information output from the recognition unit 71 and the estimated physical property information output from the learning unit 72 and the estimation unit 73, respectively, using weighting coefficients. The determination unit 74 then transmits the calculated final detected physical property information to the learning unit 72. The learning unit 72 performs learning processing based on the final detected physical property information transmitted from the determination unit 74 and the detection signal from the contact sensor 54 when that physical property information was detected.

[0052] The assignment unit 75 generates sensory information corresponding to the physical property information from the final detected physical property information output from the determination unit 74. Specifically, as shown in Figure 7, a semantic library 821 is stored in the storage device 820 controlled by the memory controller 82. The semantic library 821 stores information that shows the correspondence between multiple pieces of physical property information and multiple pieces of sensory information, for example, as shown in Figure 8. Based on the final detected physical property information output from the determination unit 74, the assignment unit 75 identifies the physical property information that is closest to the detected physical property information from among the multiple pieces of physical property information shown in Figure 8, and outputs sensory information corresponding to the identified physical property information. In this way, the assignment unit 75 corresponds to the part that adds human sensory information to the final detected physical property information output from the determination unit 74.

[0053] The language processing unit 76 converts the sensory information transmitted from the assignment unit 75 into a data format that can be transmitted to the high-speed communication network Na. The language processing unit 76 is composed of a neural network NN as shown in Figure 6. When sensory information such as "sticky" is transmitted from the assignment unit 75, the language processing unit 76 generates sensory information such as "The robot's right hand is in contact with a sticky object" so that the location where that physical property is detected can be identified. The sensory information generated by the language processing unit 76 is transmitted to the high-speed communication network Na via the data output unit 77 and the communication processing unit 78. In this embodiment, the language processing unit 76 is an example of an abstraction unit for the detection unit that abstracts sensory information.

[0054] The data compression unit 81 compresses the detection signal data of the contact sensor 54, which has been pre-processed by the pre-processing unit 70. The compressed data of the detection signal of the contact sensor 54, compressed by the data compression unit 81, is transmitted to the high-speed communication network Na via the data output unit 77 and the communication processing unit 78.

[0055] The communication processing unit 78 processes the data transmitted from the data output unit 77 to transmit it to the high-speed communication network Na, and also receives various information transmitted to the high-speed communication network Na. The various information transmitted to the high-speed communication network Na is information transmitted from other high-speed detection units Da1~Da3, Da5, high-speed drive control units Ca1~Ca3, upper-level processing units Pa1~Pa3, and intermediate-level processing units Pb1, Pb2 connected to the high-speed communication network Na. The information received by the communication processing unit 78 is transmitted to the memory controller 82 via the data receiving unit 79 and the writing processing unit 80.

[0056] The memory controller 82 controls the storage device 820 shown in Figure 7. The storage device 820 stores the semantic library 821, as well as feature data 822, related data 823, parameters 824, label data 825, and source data 826, etc. The memory controller 82 updates the various data stored in the storage device 820 based on data transmitted from, for example, the write processing unit 80. The various data stored in the storage device 820 are also used by the determination unit 74, etc.

[0057] (Configuration of the processing unit) Next, the configurations of each processing unit Pa1-Pa3, Pb1, Pb2, and Pc1-Pc3 will be described. Since each processing unit Pa1-Pa3, Pb1, Pb2, and Pc1-Pc3 has the same or similar configuration, the configuration of the right-hand intermediate processing unit Pb1, which controls the right-hand section 35a, will be described as representative below.

[0058] As shown in Figure 9, the right-hand intermediate processing unit Pb1 comprises an external interface 90, a pre-processing unit 91, a recognition unit 92, a learning unit 93, an other processing unit 94, an estimation unit 95, a determination unit 96, an assignment unit 97, a language processing unit 98, a data output unit 99, a communication processing unit 100, a data receiving unit 101, a writing processing unit 102, a data compression unit 103, a memory controller 104, an analysis unit 105, an action planning unit 106, and a self-recognition unit 107. The recognition unit 92, the learning unit 93, the other processing unit 94, the estimation unit 95, the language processing unit 98, the memory controller 104, the action planning unit 106, and the self-recognition unit 107 are connected to each other via a data-memory bus 140 so as to be able to communicate with each other. Of the elements shown in Figure 9, those with the same names as those shown in Figure 5 perform similar operations, so the differences will be explained below.

[0059] The external I / F 90 is a part to which an external device De, separate from the robot device 30, can be connected, and it acquires information transmitted from the external device De. The external device De is, for example, a unique random number generator. The information acquired by the external I / F 90 is transmitted via the pre-processing unit 91 to the recognition unit 92, the learning unit 93, the other processing unit 94, and the estimation unit 95, respectively. The other processing unit 94 is a part that performs processing on the information transmitted from the external device De, which has been pre-processed by the pre-processing unit 91, different from the processing performed by the recognition unit 92, the learning unit 93, and the estimation unit 95.

[0060] In the right-hand intermediate processing unit Pb1, for example, when sensory information transmitted from the high-speed detection unit Da4 to the high-speed communication network Na is received by the communication processing unit 100, the sensory information is input to the language processing unit 98 via the communication processing unit 100 and the data receiving unit 101, converted into a data format that can be analyzed by the analysis unit 105, and then input to the analysis unit 105. At this time, if the sensory information is, for example, "the robot's right hand is in contact with a sticky object," the analysis unit 105 extracts physical property information from that information. The extracted physical property information includes information on the location where the sensation was detected and physical property information indicating the sensory information. The physical property information indicating the sensory information includes numerical data such as hardness, elastic properties, and viscosity. The analysis unit 105 transmits the extracted physical property information as an analysis result to the memory controller 104.

[0061] The action planning unit 106 acquires the analysis results from the analysis unit 105 via the memory controller 104 and the determination unit 96. The action planning unit 106 is composed of a neural network NN as shown in Figure 6. The action planning unit 106 uses the analysis results from the analysis unit 105 as input information to generate recommended action information that the right hand unit 35a should perform. The recommended action information includes the distance, direction, and speed of movement of the right hand unit 35a. The recommended action information for the right hand unit 35a generated by the action planning unit 106 is input to the determination unit 96.

[0062] The self-recognition unit 107 acquires the analysis results from the analysis unit 105 via the memory controller 104 and the determination unit 96. The self-recognition unit 107 is composed of a neural network NN as shown in Figure 6. The self-recognition unit 107 uses the analysis results from the analysis unit 105 as input information to generate recognition information indicating the current situation in which the right hand unit 35a is located. The recognition information includes information such as whether or not an object is in contact with the right hand unit 35a, and the distance from the right hand unit 35a to any object. The recognition information of the right hand unit 35a generated by the self-recognition unit 107 is input to the determination unit 96.

[0063] The determination unit 96 generates final operation information for the right hand unit 35a based on the recommended operation information for the right hand unit 35a generated by the action planning unit 106 and the recognition information for the right hand unit 35a generated by the self-recognition unit 107. If the determination unit 96 determines, for example, that the right hand unit 35a is in contact with an object based on the recognition information for the right hand unit 35a, it increases the value of the movement speed included in the recommended operation information for the right hand unit 35a and transmits the corrected recommended operation information to the assignment unit 97 as the final recommended operation information.

[0064] The assignment unit 97 generates action command information that the right hand unit 35a should perform from the final recommended operation information for the right hand unit 35a output from the determination unit 96. Specifically, the memory device controlled by the memory controller 104 stores a semantic library 1040 that shows the correspondence between multiple operation information and multiple action command information, for example, as shown in Figure 10. Based on the final operation information of the right hand unit 35a output from the determination unit 96, the assignment unit 97 identifies the operation information that is closest to the current recommended operation information from among the multiple operation information shown in Figure 10, and outputs action command information corresponding to the identified operation information.

[0065] The language processing unit 98 converts the action command information transmitted from the assignment unit 97 into a format that can be transmitted to the high-speed communication network Na. The language processing unit 98 is composed of a neural network NN as shown in Figure 6. When the assignment unit 97 transmits action command information such as "Move the right hand 35a upward at a speed of ●●", the language processing unit 98 generates action command information such as "Immediately move the robot's right hand upward." The action command information generated by the language processing unit 98 is transmitted to the high-speed communication network Na via the data output unit 99 and the communication processing unit 100. In this embodiment, the language processing unit 98 is an example of an abstraction unit for processing that abstracts action command information.

[0066] (Configuration of the drive control unit) Next, the configurations of each drive control unit Ca1 to Ca3, Cb1, and Cb2 will be described in detail. Since each drive control unit Ca1 to Ca3, Cb1, and Cb2 has the same or similar configuration, the configuration of the high-speed drive control unit Ca2, which controls the actuator device 350a of the right hand section 35a, will be described as representative below.

[0067] As shown in Figure 11, the high-speed drive control unit Ca2 comprises a pre-processing unit 110, a recognition unit 111, a learning unit 112, an estimation unit 113, a determination unit 114, an assignment unit 115, a language processing unit 116, a data output unit 117, a communication processing unit 118, a data receiving unit 119, a writing processing unit 120, a data compression unit 121, a memory controller 122, an analysis unit 123, a control unit 124, a driver 125, and a sensor 126. The recognition unit 111, the learning unit 112, the estimation unit 113, the memory controller 122, and the language processing unit 116 are connected to each other via a data-memory bus 128 so as to be able to communicate with each other. Among the elements shown in Figure 11, those with the same names as the elements shown in Figure 5 perform similar operations, so the differences will be explained below.

[0068] In the high-speed drive control unit Ca2, for example, when action command information transmitted from the right-hand intermediate processing unit Pb1 to the high-speed communication network Na is received by the communication processing unit 118, the action command information is input to the language processing unit 116 via the communication processing unit 118 and the data receiving unit 119, converted into a data format that can be processed by the analysis unit 123, and then input to the analysis unit 123. At this time, if the action command information is "immediately move the robot's right hand upward," the analysis unit 123 generates a control signal for the actuator device 350a from that information. Specifically, the analysis unit 123 sets the drive amount and drive direction of the actuator device 350a necessary to immediately move the right hand portion 35a of the robot device 30 upward, and then transmits a control signal corresponding to the set drive amount and drive direction to the control unit 124.

[0069] The control unit 124 controls the actuator device 350a via the driver 125 based on the control signal transmitted from the analysis unit 123. This enables the actuator device 350a to perform actions according to the action command information.

[0070] Sensor 126 detects the operating state of the right hand portion 35a of the robot device 30 and outputs a detection signal corresponding to the detected operating state. In this embodiment, the right hand portion 35a corresponds to the object being worked on. Alternatively, sensor 126 may detect the operating state of actuator device 350a. The detection signal from sensor 126 is input to the recognition unit 111, learning unit 112, and estimation unit 113 via the preprocessing unit 110, and operation result information corresponding to the detection signal from sensor 126 is generated by the assignment unit 115. This operation result information is transmitted to the high-speed communication network Na via the language processing unit 116, data output unit 117, and communication processing unit 118. In this embodiment, the language processing unit 116 is an example of an abstraction unit for the drive unit that abstracts the operation result information.

[0071] The operation result information transmitted to the high-speed communication network Na by the high-speed drive control unit Ca2 of the right hand unit 35a is received by the communication processing unit 100 shown in Figure 9, for example, and input to the analysis unit 105 via the data receiving unit 101 and the language processing unit 98, where the analysis unit 105 extracts physical property information. The physical property information extracted by the analysis unit 105 is input to the action planning unit 106 via the memory controller 104 and the determination unit 114. Based on this physical property information, the action planning unit 106 can analyze whether the right hand unit 35a of the robot device 30 is operating appropriately based on the action command information it has transmitted. Furthermore, if the right hand unit 35a of the robot device 30 is not operating appropriately, the action planning unit 106 generates more appropriate recommended operation information for the right hand unit 35a and outputs the generated recommended operation information to the determination unit 96. This ensures that the right hand unit 35a of the robot device 30 operates more appropriately. In other words, feedback control of the right hand portion 35a of the robot device 30 is realized.

[0072] (Operation and effects of the control device 10 of this embodiment) As described above, the control device 10 of this embodiment controls a robot device 30 (mechanical device) which is equipped with a speaker device 43 and actuator devices 330a, 330b, 350a, and 350b. In this embodiment, the speaker device 43 and actuator devices 330a, 330b, 350a, and 350b correspond to multiple drive units. The control device 10 comprises detection units Da1 to Da5 and Db1 to Db8, processing units Pa1 to Pa3, Pb1, Pb2, and Pc1 to Pc3, and drive control units Ca1 to Ca3, Cb1, and Cb2. The detection units Da1-Da5 and Db1-Db8 convert detection signals output from each sensor 40-42 and 50-59 (detection unit sensors) installed on the robot device 30 into sensory information, verbalize (abstract) the converted sensory information, and transmit it to the communication networks Na and Nb (transmission paths). The processing units Pa1-Pa3, Pb1, Pb2, and Pc1-Pc3 analyze the sensory information transmitted from the detection units Da1-Da5 and Db1-Db8 to the communication networks Na and Nb, and based on the analysis results, generate action command information, which is information about the actions that the robot device 30 should take, and verbalize (abstract) the generated action command information and transmit it to the communication networks Na and Nb. The drive control units Ca1~Ca3, Cb1, and Cb2 analyze the action command information transmitted from the processing units Pa1~Pa3, Pb1, Pb2, and Pc1~Pc3 to the communication networks Na and Nb, and control the speaker device 43 and the actuator devices 330a, 330b, 350a, and 350b based on the analyzed action command information.

[0073] With this configuration, the detection units Da1-Da5, Db1-Db8, processing units Pa1-Pa3, Pb1, Pb2, Pc1-Pc3, and drive control units Ca1-Ca3, Cb1, Cb2 operate based on verbalized sensory information and behavioral command information. Compared to conventional control devices that operate with numerical signals or complex control signals, this configuration offers greater versatility. Furthermore, it allows for the distribution of computing resources, reducing costs and power consumption. In addition, compared to devices that operate with numerical signals or complex control signals, it reduces the amount of information that needs to be exchanged between each element, significantly reducing communication volume.

[0074] The detection units Da1~Da5 and Db1~Db8 each comprise sensors 40~42 and 50~59 (detection unit sensors), a recognition unit 71 (detection unit recognition unit), an assignment unit 75 (detection unit generation unit), a language processing unit 76 (detection unit abstraction unit), and a communication processing unit 78 (detection unit communication unit). The recognition unit 71 recognizes predetermined information detected by sensors 40~42 and 50~59 based on the detection signals from sensors 40~42 and 50~59. The assignment unit 75 generates sensory information using the predetermined information recognized by the recognition unit 71. The language processing unit 76 verbalizes (abstracts) the sensory information generated by the assignment unit 75. The communication processing unit 78 transmits the sensory information verbalized (abstracted) by the assignment unit 75 to the communication networks Na and Nb.

[0075] This configuration makes it possible to translate predetermined information detected by sensors 40-42 and 50-59 into language and transmit it to the communication networks Na and Nb.

[0076] The detection units Da1 to Da5 and Db1 to Db8 are further equipped with a learning unit 72 (learning unit for detection units). The learning unit 72 learns the relationship between the detection signals from sensors 40 to 42 and 50 to 59 and predetermined information detected by sensors 40 to 42 and 50 to 59. The assignment unit 75 generates sensory information using the learning information from the learning unit 72.

[0077] This configuration makes it possible to generate more accurate sensory information.

[0078] The processing units Pa1~Pa3, Pb1, Pb2, and Pc1~Pc3 comprise a communication processing unit 100 (communication unit for processing unit), an analysis unit 105 (analysis unit for processing unit), an action planning unit 106, a self-recognition unit 107, an assignment unit 97 (generation unit for processing unit), and a language processing unit 98 (abstraction unit for processing unit). The communication processing unit 100 receives sensory information from the communication networks Na and Nb. The analysis unit 105 analyzes the sensory information received by the communication processing unit 100. The action planning unit 106 plans the actions of the robot device 30 based on the analysis results of the sensory information by the analysis unit 105. The self-recognition unit 107 recognizes the current status of the robot device 30 based on the analysis results of the sensory information by the analysis unit 105. The assignment unit 97 generates action command information using the action plan for the robot device 30 planned by the action planning unit 106 and the current status of the robot device 30 recognized by the self-recognition unit 107. The language processing unit 98 verbalizes (abstracts) the action command information generated by the assignment unit 97. The communication processing unit 100 transmits the action command information verbalized (abstracted) by the language processing unit 98 to the communication networks Na and Nb.

[0079] This configuration makes it possible to verbalize action command information generated based on the action plan of the robot device 30 planned by the action planning unit 106 and the current status of the robot device 30 recognized by the self-recognition unit 107, and transmit it to the communication networks Na and Nb.

[0080] The processing units Pa1-Pa3, Pb1, Pb2, and Pc1-Pc3 further include an external I / F 90 (external communication unit) and a recognition unit 92 (recognition unit for processing unit). The external I / F 90 receives predetermined external input information transmitted from an external device De. The recognition unit 92 recognizes the predetermined external input information received by the external I / F 90. The assignment unit 97 further generates action command information using the predetermined external input information recognized by the recognition unit 92.

[0081] This configuration makes it possible to generate more accurate action command information.

[0082] The processing units Pa1-Pa3, Pb1, Pb2, and Pc1-Pc3 are further equipped with a learning unit 93 (learning unit for processing units). The learning unit 93 learns the relationship between external input information from an external device De and action command information. The assignment unit 97 generates action command information using the learning information from the learning unit 93.

[0083] This configuration makes it possible to generate more accurate action command information.

[0084] The drive control units Ca1 to Ca3, Cb1, and Cb2 each comprise a communication processing unit 118 (communication unit for the drive unit), an analysis unit 123 (analysis unit for the drive unit), and a control unit 124. The communication processing unit 118 receives action command information from the communication networks Na and Nb. The analysis unit 123 analyzes the action command information. The control unit 124 controls the speaker device 43 and the actuator devices 330a, 330b, 350a, and 350b based on the analysis results of the action command information by the analysis unit 123.

[0085] This configuration makes it possible to control the speaker device 43 and the actuator devices 330a, 330b, 350a, and 350b based on action command information.

[0086] The drive control units Ca1 to Ca3, Cb1, and Cb2 each include a sensor 126 (drive unit sensor), a recognition unit 111 (drive unit recognition unit), an assignment unit 115 (drive unit generation unit), and a language processing unit 116 (drive unit abstraction unit). The sensor 126 detects, for example, the operating state of actuator devices 330a, 330b, 350a, and 350b, and at least one of the operating state of the targets driven by the actuator devices 330a, 330b, 350a, and 350b. The recognition unit 111 recognizes, based on the detection signal from the sensor 126, at least one of the operating state of the actuator devices 330a, 330b, 350a, and 350b and the operating state of the targets driven by the actuator devices 330a, 330b, 350a, and 350b. The assignment unit 115 generates operation result information using at least one of the operating state of the actuator devices 330a, 330b, 350a, and 350b and the operating state of the targets driven by the recognition unit 111. The language processing unit 116 verbalizes (abstracts) the operation result information generated by the assignment unit 115. The communication processing unit 118 transmits the operation result information verbalized (abstracted) by the language processing unit 116 to the communication networks Na and Nb.

[0087] This configuration makes it possible to translate operation result information, generated based on at least one of the operating states of the actuator devices 330a, 330b, 350a, and 350b and the operating state of the driven object, into language and transmit it to the communication networks Na and Nb.

[0088] The drive control units Ca1 to Ca3, Cb1, and Cb2 are equipped with a learning unit 112 (learning unit for drive units). The learning unit 112 learns the relationship between the detection signal from the sensor 126 and at least one of the operating states of the actuator devices 330a, 330b, 350a, and 350b and the operating state of the driven object. The assignment unit 115 generates operation result information based on the learning information from the learning unit 112.

[0089] This configuration makes it possible to generate more accurate operational result information.

[0090] <Second Embodiment> Next, the control device 10 of the second embodiment will be described. The following description will focus on the differences from the control device 10 of the first embodiment.

[0091] (Overview of the control system) As shown by the dashed line in Figure 1, the control device 10 of this embodiment further comprises a plurality of abnormality detection units Dc and a plurality of abnormality drive control units Cc. The plurality of abnormality detection units Dc and the plurality of abnormality drive control units Cc are communicated to each other via a control signal line Nc. A plurality of high-speed drive control units Ca, a plurality of low-speed drive control units Cb, a plurality of upper-level processing units Pa, a plurality of intermediate-level processing units Pb, and a plurality of lower-level processing units Pc are connected to the control signal line Nc.

[0092] The abnormality detection unit Dc is the part that detects abnormalities in the mechanical device 20. The abnormality detection unit Dc has a sensor Sc for detecting abnormalities. Sensor Sc is, for example, a battery status sensor that monitors the state of the battery. The abnormality detection unit Dc transmits the abnormality information detected by sensor Sc to the control signal line Nc.

[0093] The abnormal drive control unit Cc is the part that performs fail-safe control to ensure the safety of the machine 20 or maintain the function of the machine 20 when it receives abnormal information transmitted from the abnormal detection unit Dc to the control signal line Nc, that is, when some kind of abnormality occurs in the machine 20. The abnormal drive control unit Cc controls the fail-safe drive unit Ac provided in the machine 20. The fail-safe drive unit Ac is, for example, a brake device.

[0094] (Specific configuration of machinery and control devices) As shown by the dashed line in Figure 2, the robot device 30 further includes a battery status sensor 130, an emergency stop switch 131, and a brake device 132.

[0095] The battery status sensor 130 is installed on the battery 38 of the robot device 30 and detects the status of the battery 38.

[0096] The emergency stop switch 131 is a switch that can be operated to forcibly stop the robot device 30 when some kind of abnormality occurs in the robot device 30.

[0097] The braking device 132 is a device that stops the robot device 30 from moving by generating a braking force on the soles of the robot device 30's feet. The braking device 132 is used, for example, to prevent the robot device 30 from tipping over.

[0098] As shown by the dashed line in Figure 3, the output signal of the emergency stop switch 131 is received by the higher-level processing unit Pa1. Furthermore, the higher-level processing unit Pa1 can directly control the brake device 132 by transmitting a control signal to the brake device 132 via the control signal line Nc.

[0099] The higher-level processing unit Pa3 receives the detection signal from the battery status sensor 130. Based on the detection signal from the battery status sensor 130, the higher-level processing unit Pa3 monitors the status of the battery 38, and if any abnormality occurs in the battery 38, it transmits the abnormality information to the control signal line Nc.

[0100] As shown by the dashed line in Figure 4, the first lower processing unit Pc1 can directly control the brake device 132 by transmitting a control signal to the brake device 132 via the control signal line Nc.

[0101] (Configuration of the processing unit) Next, we will describe the configuration of each processing unit Pa1-Pa3, Pb1, Pb2, and Pc1-Pc3. Since each processing unit Pa1-Pa3, Pb1, Pb2, and Pc1-Pc3 has the same or similar configuration, we will describe the configuration of the higher-level processing unit Pa1 as a representative example below.

[0102] As shown in Figure 12, the higher-level processing unit Pa1 further comprises an error processing unit 108 and an error estimation unit 109.

[0103] The abnormality processing unit 108 receives a notification from the emergency stop switch 131 via the control signal line Nc, indicating that the emergency stop switch 131 has been turned on, for example, when the emergency stop switch 131 is turned on. The abnormality processing unit 108 also receives a notification from the higher-level processing unit Pa3 via the control signal line Nc, indicating that an abnormality has occurred in the battery 38. The abnormality processing unit 108 then transmits these notifications to the abnormality estimation unit 109.

[0104] When the abnormality estimation unit 109 receives a notification indicating that the emergency stop switch 131 has been turned ON, or when it receives a notification indicating that an abnormality has occurred in the battery 38, it generates an emergency stop signal to forcibly stop each part of the robot device 30, and transmits the generated emergency stop signal to the control signal line Nc via the abnormality processing unit 108. When this emergency stop signal is transmitted to the brake device 132 via the control signal line Nc, the brake device 132 is activated and the robot device 30 comes to an emergency stop.

[0105] (Configuration of the drive control unit) Next, the configurations of each drive control unit Ca1 to Ca3, Cb1, and Cb2 will be described in detail. Since each drive control unit Ca1 to Ca3, Cb1, and Cb2 has the same or similar configuration, the configuration of the high-speed drive control unit Ca2, which controls the actuator device 350a of the right hand section 35a, will be described as representative below.

[0106] As shown in Figure 13, the high-speed drive control unit Ca2 is further provided with an abnormality processing unit 127. When the abnormality processing unit 127 receives an emergency stop signal via the control signal line Nc, it instructs the control unit 124 to forcibly interrupt and execute a process to emergency stop the actuator device 350a. Based on this instruction from the abnormality processing unit 127, the control unit 124 forcibly stops the actuator device 350a.

[0107] The same process described above is carried out in the other drive control units Ca1, Ca3, Cb1, and Cb2, which also forces the speaker device 43 and actuator devices 330a, 330b, and 350b to stop.

[0108] (Operation and effects of the control device 10 of this embodiment) As described above, the control device 10 of this embodiment is further equipped with a battery status sensor 130 and an emergency stop switch 131 (abnormality detection unit) for detecting abnormalities in the robot device 30. When an abnormality in the robot device 30 is detected by the battery status sensor 130 or the emergency stop switch 131, each drive control unit Ca1 to Ca3, Cb1, Cb2 forcibly stops the speaker device 43 and the actuator devices 330a, 330b, 350a, and 350b.

[0109] This configuration makes it possible to operate the robotic device 30 more safely.

[0110] <Third Embodiment> Next, the control device 10 of the third embodiment will be described. The following description will focus on the differences from the control device 10 of the first embodiment.

[0111] (Control device configuration) As shown in Figure 14, in the control device 10 of this embodiment, the language processing unit 76 of the high-speed detection unit Da4 has a first neural network NN10. The first neural network NN10 is a neural network capable of converting sensory information output from the assignment unit 75 of the high-speed detection unit Da4 into abstracted sensory information. The language processing unit 76 transmits the calculation results of each intermediate layer 2001 when the sensory information output from the assignment unit 75 of the high-speed detection unit Da4 is input to each input layer 2000, and the configuration information of the first neural network NN10 when the calculation results are obtained, to the high-speed communication network Na via the data output unit 77 and the communication processing unit 78.

[0112] The configuration information of the first neural network NN10 includes information on the connection weights between nodes in each layer, for example, the connection weights between nodes in multiple hidden layers 2001 and multiple output layers 2002. 11 ~wa 1n ,···,wa m1 ~wamn The following information is included. Note that m and n are arbitrary natural numbers. In addition, the configuration information of the first neural network NN10 includes information on the connections between nodes in each layer, for example, information indicating which of the multiple output layers 2002 each of the multiple hidden layers 2001 is connected to.

[0113] As shown in Figure 14, the language processing unit 98 of the intermediate processing unit Pb1 similarly has a first neural network NN10. When the language processing unit 98 obtains configuration information of the first neural network NN10 from the high-speed communication network Na via the communication processing unit 100 and the data receiving unit 101, it reflects that configuration information in its own first neural network NN10. As a result, the first neural network NN10 of the language processing unit 76 of the high-speed detection unit Da4 and the first neural network NN10 of the language processing unit 98 of the intermediate processing unit Pb1 have the same configuration.

[0114] Furthermore, after the language processing unit 98 shares the configuration of the first neural network NN10 with the language processing unit 76 of the high-speed detection unit Da4, it obtains the calculation results of each intermediate layer 2001 from the high-speed communication network Na via the communication processing unit 100 and the data receiving unit 101. By reflecting the calculation results of each intermediate layer 2001 into its own first neural network NN10, it obtains abstracted sensory information from each output layer 2002.

[0115] On the other hand, as shown in Figure 15, in the control device 10 of this embodiment, the language processing unit 98 of the intermediate processing unit Pb1 has a second neural network NN20. The second neural network NN20 is a neural network capable of converting the action command information output from the assignment unit 97 of the intermediate processing unit Pb1 into abstracted action command information. The language processing unit 98 transmits the calculation results of each intermediate layer 3001 when the action command information output from the assignment unit 97 of the intermediate processing unit Pb1 is input to each input layer 3000, and the configuration information of the second neural network NN20 when the calculation results are obtained, to the high-speed communication network Na via the data output unit 99 and the communication processing unit 100.

[0116] The configuration information of the second neural network NN20 includes information on the connection weights between nodes in each layer, for example, the connection weights wb between each node of multiple hidden layers 3001 and multiple output layers 3002. 11 ~wb 1q ,···,wb p1 ~wb pq The following information is included. Note that p and q are arbitrary natural numbers. In addition, the configuration information of the second neural network NN20 includes information on the connections between nodes in each layer, for example, information indicating which of the multiple output layers 3002 each of the multiple hidden layers 3001 is connected to.

[0117] As shown in Figure 15, the language processing unit 116 of the high-speed drive control unit Ca2 also has a second neural network NN20. When the language processing unit 116 obtains the configuration information of the second neural network NN20 transmitted from the intermediate processing unit Pb1 via the communication processing unit 118 and the data receiving unit 119 from the high-speed communication network Na, it reflects that configuration information in its own second neural network NN20. As a result, the second neural network NN20 of the language processing unit 98 of the intermediate processing unit Pb1 and the second neural network NN20 of the language processing unit 116 of the high-speed drive control unit Ca2 have the same configuration.

[0118] Furthermore, after the language processing unit 116 shares the configuration of the second neural network NN20 with the language processing unit 98 of the intermediate processing unit Pb1, it obtains the calculation results of each intermediate layer 3001 transmitted from the intermediate processing unit Pb1 via the communication processing unit 118 and the data receiving unit 119 from the high-speed communication network Na. By reflecting these calculation results of each intermediate layer 3001 into its own second neural network NN20, it obtains abstracted action command information from each output layer 3002.

[0119] Furthermore, the other detection units Da1~Da3, Da5, Db1~Db8 may have the same or similar configuration as the high-speed detection unit Da4. Also, the other processing units Pa1~Pa3, Pb2, Pc1~Pc3 may have the same or similar configuration as the intermediate processing unit Pb1. In addition, the other drive control units Ca1, Ca3, Cb1, Cb2 may have the same or similar configuration as the high-speed drive control unit Ca2.

[0120] (Operation and effects of the control device 10 of this embodiment) As described above, the high-speed detection unit Da4 of this embodiment has an input layer 2000 and an intermediate layer 2001 of the first neural network NN10 for abstracting sensory information, and transmits the calculation result of the intermediate layer 2001 when sensory information is input to the input layer 2000 to the high-speed communication network Na (transmission path). The intermediate processing unit Pb1 has an intermediate layer 2001 and an output layer 2002 of the first neural network NN10, and obtains abstracted sensory information from the output layer 2002 using the calculation result of the intermediate layer 2001 transmitted to the high-speed communication network Na by the high-speed detection unit Da4.

[0121] Even with this configuration, it becomes possible to easily acquire abstracted sensory information.

[0122] In this embodiment, the intermediate processing unit Pb1 has an input layer 3000 and an intermediate layer 3001 of the second neural network NN20 for abstracting action command information, and transmits the calculation result of the intermediate layer 3001 when action command information is input to the input layer 3000 to the high-speed communication network Na (transmission path). The high-speed drive control unit Ca2 has an intermediate layer 3001 and an output layer 3002 of the second neural network NN20, and obtains abstracted action command information from the output layer 3002 using the calculation result of the intermediate layer 3001 transmitted to the high-speed communication network Na by the intermediate processing unit Pb1.

[0123] Even with this configuration, it is possible to obtain abstracted behavioral command information.

[0124] <Other Embodiments> This disclosure is not limited to the specific examples given above.

[0125] For example, the configuration of the control device 10 described above is not limited to humanoid robot devices 30, but can also be applied to industrial robots and the like. Furthermore, the configuration of the control device 10 described above can be applied to any mechanical device, not just robots.

[0126] The language processing units 76, 98, and 116 may, instead of performing the process of translating various information into language, perform processes such as compressing various information or converting various information into symbols.

[0127] Even the above-mentioned examples, with appropriate design modifications by those skilled in the art, are included within the scope of this disclosure, as long as they possess the features of this disclosure. The elements, their arrangement, conditions, shapes, etc., of each of the above-mentioned examples are not limited to those exemplified and can be modified as appropriate. The elements of each of the above-mentioned examples can be combined in different ways as appropriate, as long as no technical inconsistencies arise.

[0128] The neural network NN in each embodiment is not limited to having a single intermediate layer 1001, but may have multiple intermediate layers. For example, if the neural network NN is a convolutional neural network (CNN), it may have convolutional layers, pooling layers, and fully connected layers as intermediate layers. The same applies to the first neural network NN10 and the second neural network NN20 in the third embodiment. [Explanation of symbols]

[0129] Aa: High-speed drive unit, Ab: Low-speed drive unit, Ca, Ca1~Ca3: High-speed drive control unit, Cb, Cb1, Cb2: Low-speed drive control unit, Da, Da1~Da5: High-speed detection unit, Db, Db1~Db8: Low-speed detection unit, Na: High-speed side communication network (transmission path), Nb: Low-speed side communication network (transmission path), Pa, Pa1~Pa3: Upper-level processing unit, Pb, Pb1, Pb2: Middle-level processing unit, Pc, Pc1~Pc3: Lower-level processing unit Unit, Sa, Sb: Sensor (Sensor for detection unit), 10: Control device, 20: Mechanical device, 30: Robot device (Mechanical device), 40: Camera sensor (Sensor for detection unit), 41: Distance sensor (Sensor for detection unit), 42: Voice sensor (Sensor for detection unit), 43: Speaker device (Drive unit), 50~59: Contact sensor (Sensor for detection unit), 71: Recognition unit (Recognition unit for detection unit), 72: Learning unit (Learning unit for detection unit), 75: 76: Assignment Unit (Generation Unit for Detection Unit), 78: Language Processing Unit (Abstraction Unit for Detection Unit), 90: Communication Processing Unit (Communication Unit for Detection Unit), 92: Recognition Unit (Recognition Unit for Processing Unit), 93: Learning Unit (Learning Unit for Processing Unit), 97: Assignment Unit (Generation Unit for Processing Unit), 98: Language Processing Unit (Abstraction Unit for Processing Unit), 100: Communication Processing Unit (Communication Unit for Processing Unit), 105: Analysis Unit (Analysis Unit for Processing Unit), 106: Action Plan Units: 107: Self-recognition unit, 111: Recognition unit (recognition unit for drive unit), 112: Learning unit (learning unit for drive unit), 115: Assignment unit (generation unit for drive unit), 116: Language processing unit (abstraction unit for drive unit), 118: Communication processing unit (communication unit for drive unit), 123: Analysis unit (analysis unit for drive unit), 124: Control unit, 126: Sensor (sensor for drive unit), 330a, 330b, 350a, 350b: Actuator device (drive unit).

Claims

1. A control device for controlling a mechanical device equipped with multiple drive units, A detection unit that converts a detection signal output from a detection unit sensor into sensory information when the detection unit sensor provided in the machine device detects predetermined information, A processing unit analyzes the sensory information converted by the detection unit and generates action command information, which is information about the action that the machine should perform, based on the results of the analysis of the sensory information. A drive control unit that controls the drive unit based on the action command information generated by the processing unit, The system comprises a transmission line connecting the detection unit, the processing unit, and the drive control unit, The detection unit has an input layer and an intermediate layer of a first neural network for abstracting the sensory information, and transmits the calculation result of the intermediate layer when the sensory information is input to the input layer, and the information of the connection weights between nodes of the first neural network to the transmission path. The processing unit has the intermediate layer and output layer of the first neural network, and uses the calculation results of the intermediate layer and the information on the connection weights between nodes of the first neural network transmitted to the transmission path by the detection unit to acquire abstracted sensory information from the output layer, analyzes the acquired abstracted sensory information, and generates the action command information based on the analysis results of the abstracted sensory information. Control device.

2. The detection unit is The aforementioned sensor for the detection unit, A detection unit recognition unit recognizes the predetermined information detected by the detection unit sensor based on the detection signal of the detection unit sensor, A generation unit for the detection unit generates the sensory information using the predetermined information recognized by the recognition unit for the detection unit, The detection unit abstraction unit has an input layer and an intermediate layer of the first neural network, and obtains the calculation result of the intermediate layer when the sensory information generated by the detection unit generation unit is input to the input layer, The system includes a detection unit communication unit that transmits the calculation results of the intermediate layer and information on the connection weights between nodes of the first neural network to the transmission path. The control device according to claim 1.

3. The detection unit further comprises a learning unit for the detection unit that learns the relationship between the detection signal from the sensor for the detection unit and the predetermined information detected by the sensor for the detection unit. The generation unit for the detection unit further uses the learning information from the learning unit for the detection unit to generate the sensory information. The control device according to claim 2.

4. The aforementioned recognition unit for the detection unit and the aforementioned learning unit for the detection unit are configured using a neural network. The control device according to claim 3.

5. A control device for controlling a mechanical device provided with a plurality of drive units, A detection unit that converts a detection signal output from a detection unit sensor into sensory information when the detection unit sensor provided in the machine device detects predetermined information, A processing unit analyzes the sensory information converted by the detection unit and generates action command information, which is information about the action that the machine should perform, based on the results of the analysis of the sensory information. A drive control unit that controls the drive unit based on the action command information generated by the processing unit, The system comprises a transmission line connecting the detection unit, the processing unit, and the drive control unit, The processing unit has an input layer and an intermediate layer of a second neural network for abstracting the action command information, and transmits the calculation result of the intermediate layer when the action command information is input to the input layer, and the information of the connection weights between nodes of the second neural network to the transmission path. The drive control unit has the intermediate layer and output layer of the second neural network, and obtains abstracted action command information from the output layer using the calculation results of the intermediate layer and the information of the connection weights between nodes of the second neural network transmitted to the transmission path by the processing unit, and controls the drive unit based on the obtained abstracted action command information. Control device.

6. The aforementioned processing unit, A communication unit for processing that receives the sensory information from the transmission line, A processing unit analysis unit that analyzes the sensory information received by the aforementioned processing unit communication unit, Based on the analysis results of the sensory information by the processing unit analysis unit, an action planning unit plans the actions of the machine device, Based on the analysis results of the sensory information by the processing unit analysis unit, a self-recognition unit recognizes the current status of the machine device, A processing unit for generating action command information, which generates the action command information using the action plan of the machine device planned by the action planning unit and the current status of the machine device recognized by the self-recognition unit, The second neural network has an input layer and an intermediate layer, and includes an abstraction unit for processing that obtains the calculation result of the intermediate layer when the action command information generated by the processing unit for processing is input to the input layer, The communication unit for the processing unit transmits the calculation results of the intermediate layer obtained by the abstraction unit for the processing unit, and the information of the connection weights between nodes of the second neural network, to the transmission path. The control device according to claim 5.

7. The aforementioned processing unit, An external communication unit that receives external input information transmitted from an external device, The system further comprises a recognition unit for processing that recognizes the external input information received by the external communication unit, The processing unit generation unit further uses the external input information recognized by the processing unit recognition unit to generate the action command information. The control device according to claim 6.

8. The processing unit further comprises a learning unit for the processing unit that learns the relationship between the external input information and the action command information. The processing unit generates the action command information using the learning information from the processing unit. The control device according to claim 7.

9. The aforementioned recognition unit for the processing unit and the aforementioned learning unit for the processing unit are configured using a neural network. The control device according to claim 8.

10. The drive control unit, A communication unit for the drive unit that receives the action command information from the transmission line, An analysis unit for the drive unit that analyzes the aforementioned action command information, The drive unit comprises a control unit that controls the drive unit based on the analysis results of the action command information by the drive unit analysis unit. The control device according to claim 1.

11. The drive control unit is A communication unit for the drive unit that receives the calculation results of the intermediate layer of the second neural network and information on the connection weights between nodes of the second neural network from the transmission path, A drive unit analysis unit that analyzes information received by the drive unit communication unit, The drive unit comprises a control unit that controls the drive unit based on the analysis results of the received information by the drive unit analysis unit. The control device according to claim 5.

12. The drive control unit, A sensor for the drive unit that detects at least one of the operating state of the drive unit and the operating state of the object moved by the drive unit, A drive unit recognition unit recognizes at least one of the operation of the drive unit detected by the drive unit sensor and the operation of the object to be worked on, based on the detection signal of the drive unit sensor. A drive unit generation unit generates operation result information using at least one of the operation of the drive unit and the operation of the object recognized by the drive unit recognition unit, The system further comprises an abstraction unit for the drive unit that abstracts the operation result information generated by the drive unit generation unit, The communication unit for the drive unit transmits the operation result information abstracted by the abstraction unit for the drive unit to the transmission path. The control device according to claim 10 or 11.

13. The control unit further includes a drive unit learning unit that learns the relationship between the detection signal of the drive unit sensor and at least one of the operation of the drive unit and the operation of the object being worked on. The drive unit generation unit further uses the learning information from the drive unit learning unit to generate the operation result information. The control device according to claim 12.

14. The aforementioned recognition unit for the drive unit, the abstraction unit for the drive unit, and the learning unit for the drive unit are configured by a neural network. The control device according to claim 13.

15. The processing unit comprises a plurality of lower-level processing units, a plurality of intermediate-level processing units whose processing takes precedence over the lower-level processing units, and a plurality of upper-level processing units whose processing takes precedence over the lower-level processing units and the intermediate-level processing units. The control device according to claim 1 or 5.

16. The detection unit comprises a plurality of low-speed detection units and a plurality of high-speed detection units that operate at a higher speed than the low-speed detection units. The control device according to claim 1 or 5.

17. The drive control unit comprises a plurality of low-speed drive control units and a plurality of high-speed drive control units that operate at a higher speed than the low-speed drive control units. The control device according to claim 1 or 5.

18. The machine device is further equipped with an abnormality detection unit for detecting abnormalities, The drive control unit forcibly stops the drive unit when the abnormality detection unit detects an abnormality in the mechanical device. The control device according to claim 1 or 5.

Images