Model generation system for prosthetic leg control, prosthetic leg control command generation system, knee joint, method for generating a model for prosthetic leg control, and program for generating a model for prosthetic leg control.

Abstract

This technology provides the ability to quickly generate control command information for controlling a prosthetic leg according to the required movement. [Solution] The prosthetic leg control model generation system comprises an information acquisition unit 102 that acquires control-related information 150 related to the motion control of an electronically controlled prosthetic leg, and a model generation unit 104 that generates a generation model 106 that takes a motion request for the prosthetic leg as input and outputs control command information for the prosthetic leg, based on the information acquired by the information acquisition unit 102.

Description

Technical Field

[0004] ,

[0006] , , ,

[0005] , , ,

[0001] The present invention relates to a prosthesis control model generation system, a prosthesis control command generation system, a knee joint, a prosthesis control model generation method, and a prosthesis control model generation program.

Background Art

[0002] Some electronically controlled prostheses are prepared with a plurality of motion modes in advance in order to enable different operation controls according to the usage mode of the prosthesis.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] There has been an increasing demand for electronically controlled prostheses to be able to reproduce various motions desired by the user. However, there are cases where the motions desired by the user cannot be reproduced with the pre-prepared modes. In such cases, for example, it is conceivable to cope with the problem by updating the firmware of the prosthesis, but it may take a long time to cope with it.

[0005] In view of the above problems, an object of the present invention is to provide a technology capable of quickly generating control command information for controlling a prosthesis according to a required motion.

Means for Solving the Problems

[0006] To solve the above problems, a model generation system for controlling a prosthetic leg according to one aspect of the present invention comprises an information acquisition unit that acquires control-related information relating to the motion control of an electronically controlled prosthetic leg, and a model generation unit that takes motion requests for the prosthetic leg as input and outputs control command information for the prosthetic leg, and generates a generation model based on the information acquired by the information acquisition unit.

[0007] A prosthetic leg control command generation system according to one aspect of the present invention comprises a server, a communication terminal capable of communicating with the server, and a prosthetic leg capable of communicating with the communication terminal, wherein the server comprises an information acquisition unit that acquires control-related information relating to the operation control of an electronically controlled prosthetic leg, and a model generation unit that generates a generation model that takes an operation request of the prosthetic leg as input and outputs control command information of the prosthetic leg based on the information acquired by the information acquisition unit, the communication terminal comprises a request acquisition unit that acquires an operation request of the prosthetic leg from a user, the server further comprises a control command output unit that inputs the operation request acquired by the request acquisition unit to the generation model and outputs the control command information output from the generation model as generated control command information, and the prosthetic leg comprises a control unit that controls the operation of the prosthetic leg in accordance with the control command information and a setting unit that causes the generated control command information to be set in the control unit.

[0008] A knee joint according to one aspect of the present invention is an electronically controlled knee joint comprising a control unit that controls rotational resistance around the knee axis in accordance with control command information, comprising: an information acquisition unit that acquires control-related information relating to the operation control of the knee joint; a model generation unit that generates a generation model that takes an operation request of the knee joint as input and outputs control command information of the knee joint based on the information acquired by the information acquisition unit; a request acquisition unit that acquires an operation request of the knee joint from a user; a control command output unit that inputs the operation request acquired by the request acquisition unit to the generation model and outputs the control command information output from the generation model as generated control command information; and a setting unit that causes the control unit to set the generated control command information.

[0009] A method for generating a prosthetic leg control model according to one aspect of the present invention is a method for generating a prosthetic leg control model performed by a computer, and includes an acquisition step of acquiring control-related information relating to the motion control of an electronically controlled prosthetic leg, and a generation step of generating a generation model that takes motion requests of the prosthetic leg as input and outputs control command information of the prosthetic leg, based on the information acquired in the acquisition step.

[0010] A prosthetic leg control model generation program according to one aspect of the present invention is a program that causes a computer to execute an acquisition step of acquiring control-related information relating to the motion control of an electronically controlled prosthetic leg, and a generation step of generating a generation model that takes motion requests of the prosthetic leg as input and outputs control command information of the prosthetic leg, based on the information acquired in the acquisition step.

[0011] Furthermore, any combination of the above, or any substitution of the components or expressions of the present invention between methods, apparatus, programs, temporary or non-temporary storage media recording programs, systems, etc., are also valid embodiments of the present invention. [Effects of the Invention]

[0012] According to the present invention, it is possible to provide a technology that can quickly generate control command information for controlling a prosthetic leg in accordance with the required movement. [Brief explanation of the drawing]

[0013] [Figure 1] This figure shows a schematic configuration of a prosthetic leg control command generation system according to one embodiment. [Figure 2] Figure 1 schematically shows the functional blocks of the prosthetic leg control command generation system. [Figure 3] This flowchart shows an example of processing performed by a model generation system for prosthetic limb control. [Figure 4] Figure 1 is a flowchart showing an example of processing by the prosthetic leg control command generation system. [Modes for carrying out the invention]

[0014] Among the embodiments disclosed herein, those composed of multiple objects may be integrated, and conversely, those composed of a single object may be divided into multiple objects. Whether or not they are integrated, the invention can be constructed in a way that achieves its objective.

[0015] In the embodiments disclosed herein, those in which multiple functions are provided in a distributed manner may have some or all of those multiple functions integrated into a single unit, and conversely, those in which multiple functions are integrated may have some or all of those functions provided in a distributed manner. Whether the functions are integrated or distributed, the configuration should be such that the objective of the invention can be achieved.

[0016] Furthermore, separate components that share common characteristics are distinguished by adding "1st," "2nd," etc., to the beginning of their names, and these are omitted when referring to them collectively. In addition, terms containing ordinal numbers such as "1st," "2nd," etc., are used to describe various components, but these terms are used solely for the purpose of distinguishing one component from others, and do not limit the components themselves.

[0017] The present disclosure will be described below with reference to the drawings, based on preferred embodiments. In embodiments and modifications, the same or equivalent components and members will be denoted by the same reference numerals, and redundant descriptions will be omitted as appropriate. In addition, the dimensions of members in each drawing will be enlarged or reduced as appropriate for ease of understanding. Furthermore, some members that are not important for explaining the embodiments will be omitted from the drawings.

[0018] An embodiment will be described below with reference to the drawings. Figure 1 is a diagram showing a schematic configuration of a prosthetic leg control command generation system 1 according to one embodiment. The prosthetic leg control command generation system 1 comprises a server 2, a communication terminal 4, and an electronically controlled prosthetic leg 10.

[0019] Server 2 and communication terminal 4 can communicate with each other via first wireless communication 6. Communication terminal 4 and prosthetic leg 10 can communicate with each other via second wireless communication 8. Communication terminal 4 is a communication terminal used by the user of prosthetic leg 10, and as an example, it is a portable terminal such as a smartphone or a tablet terminal. Any wireless communication technology can be used for first wireless communication 6 and second wireless communication 8. First wireless communication 6 is, for example, wireless communication such as wireless LAN (Local Area Network) communication or mobile phone communication. Second wireless communication 8 is, for example, short-range wireless communication such as BLE (Bluetooth (registered trademark) Low Energy).

[0020] First, an overview of prosthetic leg 10 will be described. Prosthetic leg 10 includes a plastic socket 11 corresponding to the user's thigh, a thigh connection part 22 to which socket 11 is connected, a lower leg part 21 rotatably connected to thigh connection part 22 around a knee axis (not shown), and a foot part 12 connected to the lower end of lower leg part 21. In this specification, the "front-back direction" refers to the front-back direction from the perspective of the user of prosthetic leg 10, with the front being the left side in FIG. 1 and the back being the right side in FIG. 1. Similarly, in this specification, the "left-right direction" refers to the left-right direction from the perspective of the user of prosthetic leg 10, with the left being the front side of the paper in FIG. 1 and the right being the back side of the paper in FIG. 1.

[0021] Thigh connection part 22 and lower leg part 21 rotate relative to each other around a knee axis perpendicular to the paper surface of FIG. 1 provided at their connection part, thereby flexing and extending knee joint 20 corresponding to the knee joint. Also, foot part 12 is formed of an elastic member, and the relative posture when the load from the ground is small, such as when not on the ground or standing upright, is kept constant by elasticity. Also, when the load from the ground is large, such as during walking, the elastic member elastically deforms to generate a propulsive force for kicking the ground.

[0022] The knee joint 20 includes a lower leg portion 21 formed by a high-strength frame, a thigh connection portion 22 that is connected to a socket 11 corresponding to the user's thigh and is rotatably connected to the lower leg portion 21 around the knee axis, a cylinder 30 that restricts or permits the rotational movement around the knee axis, that is, the flexion and extension movement of the knee joint 20, and a control mechanism 40 that drives the cylinder 30.

[0023] The amount of expansion and contraction of the cylinder 30 and the knee angle, which is the rotational angle of the knee joint 20 around the knee axis, almost correspond one-to-one. A knee angle sensor 60 that detects the knee angle of the knee joint 20 by measuring the amount of expansion and contraction of the cylinder 30 is provided near the cylinder 30 and the thigh connection portion 22. The knee angle detected by the knee angle sensor 60 is used by the control device 50. The knee angle sensor 60 can be constituted by any sensor capable of measuring the amount of expansion and contraction of the cylinder 30. For example, it can be constituted by a hall element capable of detecting the position of a magnet embedded in a piston rod 34 that moves as the cylinder 30 expands and contracts. The knee angle is the angle formed by the axis of the socket 11 corresponding to the user's thigh and the axis of the lower leg portion 21 around the knee axis. For example, when the user of the prosthetic leg 10 stands upright and the axes of the socket 11 and the lower leg portion 21 are in a straight line, the knee angle is 0°. Also, when the user of the prosthetic leg 10 sits down and the axis of the lower leg portion 21 remains in the vertical direction while the axis of the socket 11 changes to the horizontal direction, the knee angle becomes 90°.

[0024] The method for detecting the knee angle is not limited to the above method. For example, as a method for detecting the knee angle, a method of directly measuring the knee angle with a rotary encoder installed on the knee axis of the knee joint 20 may be used, or a method of mechanically converting the rotational movement around the knee axis of the knee joint 20 into the amount of lateral slide at a predetermined location of the knee joint 20, measuring the slide amount with a hall element, and converting it into an angle may also be used.

[0025] The inertial sensor 75, located on the lower leg portion 21, detects the posture and motion of the lower leg portion 21 by measuring the velocity (angular velocity) and / or acceleration (angular acceleration) of the three axes that govern the motion of the lower leg portion 21 in the translational and / or rotational directions. Of the posture of the lower leg portion 21 detected by the inertial sensor 75, the lower leg angle, which is the inclination angle that the axis of the lower leg portion 21 makes with respect to the vertical line, is used to control the knee joint 20. The lower leg angle and the thigh angle, which will be described later, can be detected in two directions, the anterior-posterior direction and the lateral direction, and can be used to control the knee joint 20 in both directions. The inertial sensor 75 can be installed at any location on the lower leg portion 21, but for example, it is mounted together with the control device 50 on a control board installed on the outer circumference of the cylinder 30.

[0026] The thigh angle, which is the inclination angle of the axis of the socket 11 corresponding to the user's thigh with respect to the vertical line, can be calculated from the knee angle acquired by the knee angle sensor 60 and the lower leg angle acquired by the inertial sensor 75. That is, the lower leg angle is the inclination of the lower leg 21 from the vertical line, and the knee angle is the inclination of the axis of the socket 11 from the axis of the lower leg 21, so by adding the two together, the thigh angle, which is the inclination of the axis of the socket 11 from the vertical line, can be calculated. Alternatively, the thigh angle may be directly measured by providing an inertial sensor on the socket 11 or the thigh connection part 22. In this case, the lower leg angle can be calculated based on the measured knee angle and thigh angle. Similarly, even if the knee angle sensor 60 is not provided, the knee angle can be calculated by measuring the thigh angle and lower leg angle. In other words, the thigh angle, lower leg angle, and knee angle are related such that if information on two of these angles is obtained, information on the other angle can also be obtained.

[0027] In addition to the knee angle sensor 60 and the inertia sensor 75, a load sensor 70 for detecting the load applied between the lower leg portion 21 and the foot portion 12 may be provided at the lower end of the lower leg portion 21. The load sensor may also be provided at the thigh connection portion 22 or at the knee portion between the thigh connection portion 22 and the lower leg portion 21. Furthermore, a temperature sensor 80 for measuring the temperature of the oil inside the cylinder 30 may be attached to the outer or inner wall of the cylinder 30. The measurement information from each of these sensors is used by the control device 50.

[0028] A vibrator 85 may be provided in the thigh connection portion 22 or the lower leg portion 21. The vibrator 85 provides vibrational notifications and warnings to the user wearing the prosthetic leg 10 and is controlled by the control device 50.

[0029] The control device 50 controls the control mechanism 40 based on measurement information from various sensors such as the knee angle sensor 60, load sensor 70, inertia sensor 75, and temperature sensor 80, thereby controlling the resistance to the extension and retraction movement of the cylinder 30, i.e., the rotational resistance of the knee axis during the flexion movement of the knee joint 20. A battery 55 that supplies power to each part of the knee joint 20 is connected to the control device 50. In Figure 1, the control mechanism 40, control device 50, and battery 55 are shown outside the knee joint 20, but they may be provided inside the lower leg portion 21 as components of the knee joint 20.

[0030] Cylinder 30 is a hydraulic cylinder that uses oil as a working fluid to generate resistance, thereby limiting or allowing the flexion or extension movement of the knee joint 20. Cylinder 30 is supported by an upper support point 31 located near the knee axis 23, which rotatably connects the thigh joint 22 and the lower leg portion 21, and a lower support point 32 connected to a part of the lower leg portion 21, and is extendable and retractable between the two support points. In the contraction process, when the cylinder length is reduced, the lower leg portion 21 performs a flexion movement in which it rotates counterclockwise around the knee axis relative to the thigh joint 22 in the direction shown in Figure 1. In the extension process, when the cylinder length is increased, the lower leg portion 21 performs an extension movement in which it rotates clockwise around the knee axis relative to the thigh joint 22 in the direction shown in Figure 1. Here, cylinder length refers to the length between the upper support point 31 and the lower support point 32 of cylinder 30.

[0031] Figure 2 is a schematic diagram showing the functional blocks of the prosthetic leg control command generation system 1. Note that Figure 2 shows only the control device 50 of the components of the prosthetic leg 10 included in the prosthetic leg control command generation system 1. Server 2 is equipped with a first communication unit 100. Communication terminal 4 is equipped with a second communication unit 160. Control device 50 is equipped with a third communication unit 180. Each of the first communication unit 100, the second communication unit 160, and the third communication unit 180 includes one or more wireless communication modules. The first communication unit 100 of server 2 and the second communication unit 160 of communication terminal 4 can communicate with each other via the first wireless communication 6. The second communication unit 160 of communication terminal 4 and the third communication unit 180 of control device 50 can communicate with each other via the second wireless communication 8.

[0032] Each functional block shown in Figure 2 can be implemented in hardware terms using electronic elements, electronic circuits, and mechanical parts such as the computer's processor, CPU, and memory, and in software terms using computer programs, etc. However, here we depict functional blocks that are realized through the coordination of these elements. Therefore, it will be understood by those skilled in the art that these functional blocks can be implemented in various ways through combinations of hardware and software.

[0033] As will be explained in detail later, the prosthetic leg control command generation system 1 has the function of receiving an operation request for the prosthetic leg 10 from the user of the prosthetic leg 10, outputting control command information based on the operation request using a pre-generated generation model 106, and operating the prosthetic leg 10 according to the control command information. In addition, the prosthetic leg control command generation system 1 also functions as a model generation system for prosthetic leg control that generates the generation model 106. Below, we will first describe the configuration and function of the prosthetic leg control command generation system 1 as a model generation system for prosthetic leg control.

[0034] Server 2 comprises an information acquisition unit 102 and a model generation unit 104. The information acquisition unit 102 acquires control-related information 150, which is information related to the motion control of the prosthetic leg 10, and constraint information 152 of the prosthetic leg 10. The model generation unit 104 generates a generated model 106 based on the information acquired by the information acquisition unit 102. The generated model 106 is a model that takes motion requests of the prosthetic leg 10 as input and outputs control command information of the prosthetic leg 10.

[0035] The control-related information 150 defines the motion control permitted for the prosthetic leg 10. The control-related information 150 is information that defines at least one of the motion control and state ranges that the prosthetic leg 10 can achieve. The control-related information 150 may include sensor attribute information. The sensor attribute information includes information about the measurement area of ​​various sensors, information about the measured values ​​of various sensors, specification information that defines the measurement conditions of various sensors, etc., when the prosthetic leg 10 performs normal motion control. The control-related information 150 may also include control parameters of the prosthetic leg 10. The control-related information 150 may also include information that defines the correspondence between the sensor attribute information and the control parameters of the prosthetic leg 10. The information that defines the motion control may include information such as the time required from the input of a control command to the start of the operation, and the walking algorithm applied according to the walking cycle. The information that defines the state range may include information such as the state range related to the rotational resistance of the knee axis, the range of the knee angle, and the temperature range in which the prosthetic leg 10 can operate. The control-related information 150 includes, for example, the following information. • During the flexion movement of the knee joint 20, the rotational resistance of the knee axis can be changed from a light state to a fixed, immobile state. • During the extension movement of the knee joint 20, the rotational resistance of the knee axis can be changed from a light state to a fixed, immobile state within a range of knee angle from 0° to 30°. The knee angle sensor 60 can measure the knee angle in 1° increments from 0° to 150°. The temperature sensor 80 is capable of measuring the oil temperature in the cylinder 30 from -20°C to 100°C. - The force acting on the knee joint 20 can be measured as a moment from 0 Nm to 100 Nm, either towards the toe or heel. The load sensor 70 is capable of measuring vertical loads applied to the knee joint 20 from 0N to 1500N. • The acceleration acting on the knee joint 20 can be measured in three axial directions. • The knee joint 20 has a gyro sensor that can detect movements such as tilting and rotation of the prosthetic leg 10. • The ability to calculate the thigh angle from the tilt and knee angle of the prosthetic leg 10. • It takes 10 milliseconds for the control program of the prosthetic leg 10 to calculate the result and reflect it in the function of the prosthetic leg 10. • Use a walking algorithm when transitioning from the stance phase to the swing phase and from the swing phase to the stance phase.

[0036] Furthermore, the control-related information 150 may include at least a portion of the constraint information 152. The constraint information 152 is information that limits at least one of the range of motion control and state of the prosthetic leg 10. Even if a motion control or state is achievable by the prosthetic leg 10 based on the control-related information 150, there may be motion controls or states that should be limited by the constraint information 152. The constraint information 152 is information specified to ensure the safety of the user of the prosthetic leg 10. Specifically, the constraint information 152 includes first constraint information to avoid overloading the prosthetic leg 10 from the perspective of the structure of the prosthetic leg 10. The constraint information 152 also includes second constraint information to avoid the inability to fulfill the function of the prosthetic leg 10 from the perspective of control of the prosthetic leg 10. Furthermore, the constraint information 152 includes third constraint information to avoid from the perspective of the human body structure of the user of the prosthetic leg 10. The first constraint information is, for example, as follows: • Lock the knee joint 20 at a deep knee angle of 70° or more. This is because locking the knee joint 20 at a deep knee angle results in a large mechanical load and a large load on the prosthetic leg 10. The third constraint information is as follows, for example: - The knee joint 20 is locked in extension when the knee angle is slightly flexed during swing-leg walking. This is because the knee joint 20 has begun to extend in this state, and locking it in extension at this point makes the user more likely to fall. - The release point, which is the threshold at which the knee joint 20 is deemed to be able to flex during walking, is not changed by more than a certain value. Extreme changes in the release point are undesirable from the standpoint of the user's risk of falling, etc. • When weight is placed on the heel, the flexion resistance of the knee joint 20 should be free. This is because knee buckling could increase the risk of the user falling. • When weight is applied to the knee joint 20 while it is locked, the flexion resistance of the knee joint 20 should be released. This is because knee buckling could increase the risk of the user falling.

[0037] The information acquisition unit 102 may acquire other information. This other information may include, for example, general information about the biological movement mechanism related to the movement of the human body, or control information based on the information about the biological movement mechanism. The information acquisition unit 102 may acquire the control-related information 150, the constraint information 152, and other information from the storage unit 130 provided in the server 2, or from outside the server 2.

[0038] As described above, the model generation unit 104 generates a generated model 106 based on the information acquired by the information acquisition unit 102, namely control-related information 150, constraint information 152, and other information. The generated model 106 is a trained model that takes the operation request of the prosthetic leg 10 as input and outputs control command information of the prosthetic leg 10. The model generation unit 104 may also generate the generated model 106 using a generated AI model that has been generated through learning using information periodically acquired from various sites via the internet. The generated AI model is configured to accept, for example, natural language text (also called "natural language text") as input and generate various responses corresponding to the natural language text based on common sense.

[0039] When the model generation unit 104 generates a generation model 106 using the generation AI model, the model generation unit 104 inputs information based on control-related information 150, constraint information 152, and other information acquired by the information acquisition unit 102 as a prompt to the generation AI model. The model generation unit 104 then inputs a prompt to the generation AI model instructing it to output control command information for the prosthetic leg 10 in response to an operation request for the prosthetic leg 10 when such a request is received. This completes the generation model 106, which takes an operation request for the prosthetic leg 10 as input and outputs control command information for the prosthetic leg 10, and is configured such that the control command information satisfies constraints based on the information acquired by the information acquisition unit 102. The model generation unit 104 may also generate the generation model 106 using machine learning other than the generation AI model.

[0040] When the model generation unit 104 generates the generation model 106 using machine learning other than the generated AI model, supervised learning using predetermined training data may be used. In this case, the training data may be, for example, data based on the correspondence between measurement information from various sensors when the prosthesis 10 performs normal motion control and the control parameters of the prosthesis 10, which is included in the control-related information 150; data based on constraint information 152; or data based on information on control conditions and results in simulations of prosthesis control previously performed by the simulation unit 114 or other simulation means. The training data may also be data based on other knowledge of prosthesis walking. As knowledge of prosthesis walking, for example, the content described in "Amputation and Prosthetics," 2nd edition, Ishiyaku Publishers, Inc., January 2016, "Chapter 4 Prostheses" can be used. Examples of knowledge regarding prosthetic gait include insights into the stance phase, swing phase, skeletal muscle movement, balance and postural control, and the involvement of the nervous system in motor control in healthy individuals, as well as insights into the stance phase, swing phase, necessary control and techniques for prosthetics, balance and postural control, the involvement of the nervous system in motor control, and prosthetic fitting and adjustment in prosthetic users.

[0041] The operation request for the prosthetic leg 10 input to the generation model 106 includes information about the operation content desired by the user of the prosthetic leg 10. The operation content information may include information about operations that change over time. The control command information for the prosthetic leg 10 output from the generation model 106 is information for commanding the prosthetic leg 10 to perform operation control, and may be control parameters for performing operation control of the prosthetic leg 10, or it may be an algorithm that includes information about the control parameters. If the control command information is an algorithm, for example, it may be an algorithm that changes the control parameters to predetermined values ​​when the measurement information from various sensors of the prosthetic leg 10 satisfies predetermined conditions. The control parameters may be parameters of the rotational resistance of the knee joint 20.

[0042] Thus, the prosthetic leg control command generation system 1 can generate a generation model 106 that takes an operation request for the prosthetic leg 10 as input and outputs control command information for the prosthetic leg 10, based on information such as control-related information 150 and constraint information 152. This allows for the rapid generation of control command information for controlling the prosthetic leg according to the requested operation. Therefore, the prosthetic leg 10 can be controlled quickly according to the requested operation without having to update the firmware of the prosthetic leg 10. Furthermore, since the prosthetic leg 10 has constraints on its operation and state, it was necessary to consider these constraints when generating control command information corresponding to the requested operation. However, with this configuration, control command information can be generated with a high degree of freedom while taking these constraints into consideration.

[0043] Figure 3 is a flowchart showing an example of processing by the prosthetic leg control model generation system. As described above, the prosthetic leg control model generation system is a function of the prosthetic leg control command generation system 1.

[0044] The information acquisition unit 102 acquires control-related information 150 and constraint information 152 (S10). The information acquisition unit 102 may acquire other information as well.

[0045] The model generation unit 104 inputs the information acquired by the information acquisition unit 102 as a prompt into the generated AI model (S12). The model generation unit 104 also inputs a prompt into the generated AI model instructing it to take the operation request of the prosthetic leg 10 as input and the control command information of the prosthetic leg 10 as output (S14).

[0046] In this way, the generative AI model is trained, and when an action request for the prosthetic leg 10 is input, a generative model 106 is generated that outputs control command information for the prosthetic leg 10 (S16).

[0047] Returning to the explanation of Figure 2, the prosthetic leg control command generation system 1 further comprises the following components in addition to the configuration as a model generation system for prosthetic leg control described above. The server 2 comprises a control command output unit 108, a determination unit 110, a notification unit 112, a simulation unit 114, and a modification unit 116. The communication terminal 4 comprises a request acquisition unit 162. The control device 50 for the prosthetic leg 10 comprises a setting unit 182, a control unit 184, and a sensor information acquisition unit 186.

[0048] In the communication terminal 4, the request acquisition unit 162 acquires an action request from the user of the prosthetic leg 10. The request acquisition unit 162 may acquire natural language as the action request. Alternatively, the request acquisition unit 162 may use a generation AI function to present the user with multiple questions and generate an action request based on the user's answers. This allows the user to easily input an action request by simply intuitively inputting using natural language or by answering questions.

[0049] The communication terminal 4 functions as a request acquisition unit 162 by executing software such as a dedicated application for use in the prosthetic leg control command generation system 1 and acquiring operation requests input via any input means provided by the communication terminal 4, such as touch operation using a touchscreen or voice input using a microphone. The software executed by the communication terminal 4 may be capable of performing various functions while communicating with the server 2 and the control device 50 as appropriate via the second communication unit 160. The communication terminal 4 transmits the operation requests acquired via the request acquisition unit 162 to the server 2 via the second communication unit 160.

[0050] In server 2, the control command output unit 108 outputs control command information from the generation model 106 by inputting the operation request acquired by the request acquisition unit 162 to the generation model 106. The control command information output by the control command output unit 108 is also called generated control command information. If the operation request acquired by the request acquisition unit 162 is in natural language, the generation model 106 may interpret the operation request by processing the natural language.

[0051] The determination unit 110 determines the suitability of the generation control command information output by the control command output unit 108 based on at least one of the operation request and the constraint information 152. Specifically, the determination unit 110 determines at least one of the following: that the generation control command information does not conform to the operation request, and that the generation control command information violates the constraint information 152. The generation control command information is required to conform to the operation request and not violate the constraint information 152. However, if conformity to the operation request is prioritized, the generation model 106 may output generation control command information that violates at least a part of the constraint information 152. Also, if non-violation of the constraint information is prioritized, the generation model 106 may output generation control command information that does not conform to at least a part of the operation request. Furthermore, the generation model 106 may not be perfect, and even if none of the above circumstances apply, the generation control command information may not conform to the operation request or may violate the constraint information 152. In order to detect that the generation control command information is unsuitable, the determination unit 110 performs the above-mentioned suitability determination.

[0052] The notification unit 112, in accordance with the determination result by the determination unit 110, notifies the user of the prosthetic leg 10 if the generated control command information is unsuitable. The notification may be made using the notification means of the communication terminal 4, or it may be made using the vibrator 85 of the prosthetic leg 10. This allows the user to be informed that the control of the prosthetic leg 10 cannot be performed in response to the operation request.

[0053] The request acquisition unit 162 may acquire an operation request from the user again after receiving notification from the notification unit 112 that the generation control command information is unsuitable. In this case, the request acquisition unit 162 may use the generation AI function to present the user with additional questions and generate or modify the operation request based on the user's answers.

[0054] The simulation unit 114 uses the generated control command information output by the control command output unit 108 to perform a simulation of prosthetic leg control in software. Specifically, the simulation unit 114 appropriately inputs values ​​from various sensors that are expected when the user performs actions such as walking, and controls the operation of the prosthetic leg 10 in software. The simulation unit 114 then determines whether or not an error such as an infinite loop occurs in the program through the simulation. The notification unit 112 may also notify the user of the prosthetic leg 10 if the simulation by the simulation unit 114 determines that an error has occurred.

[0055] The modification unit 116 modifies the generation control command information based on the simulation results from the simulation unit 114, for example, when an error occurs. This allows for appropriate correction in response to program inconsistencies or other issues in the generation control command information.

[0056] Server 2 transmits the generation control command information output from the control command output unit 108 or the generation control command information modified by the modification unit 116 to the communication terminal 4 via the first communication unit 100. The communication terminal 4 transmits the generation control command information received from Server 2 to the control device 50 of the prosthetic leg 10 via the second communication unit 160.

[0057] In the control device 50 of the prosthetic leg 10, the setting unit 182 causes the control unit 184 to set the generation control command information. The setting unit 182 may also add the generation control command information to the firmware of the control device 50. The control unit 184 controls the control mechanism 40 to control the rotational resistance of the knee joint 20 around the knee axis according to the generation control command information. The sensor information acquisition unit 186 acquires measurement information from various sensors of the prosthetic leg 10. The control unit 184 may use the measurement information acquired by the sensor information acquisition unit 186 as feedback information to adjust the control of the control mechanism 40.

[0058] The prosthetic leg control command generation system 1 may, for example, receive user operations using software executed on a communication terminal 4, set names and numbers for the generated control command information, store it in a storage unit 130, or delete it according to user operations.

[0059] Multiple generation models 106 may be provided, and each of the multiple generation models 106 may be associated with a different user ID, etc. Furthermore, the server 2 may further include a model update unit 118. The model update unit 118 updates the generation models 106 based on at least one of the simulation results of prosthetic leg control using generation control command information from the simulation unit 114, and the actual control results of the prosthetic leg 10 using the generation control command information. As the actual control results of the prosthetic leg 10 using the generation control command information, for example, measurement information acquired by the sensor information acquisition unit 186 can be used. The measurement information acquired by the sensor information acquisition unit 186 may change depending on the user's movement characteristics, physical information such as weight and height, etc. The model update unit 118 may update each of the multiple generation models 106 individually for each user ID.

[0060] This allows the generated model 106 to be optimized for each user. Furthermore, the model update unit 118 can improve the accuracy of the generated model 106 by updating it based on the simulation results.

[0061] Server 2 may further include a shared unit 120. The shared unit 120 has the function of sharing generation control command information generated using one of the multiple generation models 106 with users of other generation models 106. For example, in response to a user operation received using software executed on the communication terminal 4, the generated control command information generated for that user may be registered in the shared unit 120, making it available to other users. Alternatively, in response to a user operation received using software executed on the communication terminal 4, the generated control command information of other users registered in the shared unit 120 may be selected and added to the firmware of the control device 50 of the user's prosthetic leg 10. When adding the generated control command information of other users to the firmware of the control device 50 of the prosthetic leg 10, the generated control command information may be added as is, or the information may be modified using correction values ​​for each user, such as values ​​that reflect the user's preferences or physical information.

[0062] This allows users to share generation control command information generated using a user-optimized generation model 106 among themselves.

[0063] Figure 4 is a flowchart showing an example of processing by the prosthetic leg control command generation system 1.

[0064] The request acquisition unit 162 acquires an action request from the user of the prosthetic leg 10 (S20). The action request is a natural language statement, for example, "I want to climb a stepladder to do some gardening. After I climb up the stepladder, I want my knees to stay bent only slightly, not fully bent."

[0065] The communication terminal 4 transmits the operation request acquired via the request acquisition unit 162 to the server 2 via the second communication unit 160. In the server 2, the control command output unit 108 inputs the operation request acquired by the request acquisition unit 162 to the generation model 106 (S22), and outputs control command information from the generation model 106 (S24). At this time, the generation model 106 performs natural language processing or other analysis on the natural language sentence of the operation request and decomposes it into a time-series of operation content, for example, first "climb the stepladder", then "stop when your knees are slightly bent after climbing to the top of the stepladder", and finally "descend from the stepladder". The generation model 106 may then combine control command information corresponding to each operation content and control command information that defines the conditions for switching each operation content, etc., and output it as generated control command information.

[0066] The generation control command information in this example includes, for example, the following: (1) Create a state where there is no resistance to knee flexion (free resistance). This is the state in which you can place your foot on a step ladder. (2) In addition to load detection by the load sensor 70, when landing impact is detected by the inertia sensor 75, the flexion side of the knee axis is locked. The extension side remains free-resisting. This is based on the judgment that the foot is on the stepladder. (3) When the knee is nearly fully extended and the load is removed, and acceleration is detected causing the leg to move backward, the flexion side of the knee axis is made free resistance. This is based on the judgment that the person is climbing to the next step while their foot is on the stepladder. Repeat (2) and (3) thereafter. (4) When the knee and thigh angles are nearly extended, and a certain load is detected, and this condition continues for several seconds, a slight resistance to knee axis flexion is generated, and extension and flexion lock are applied in a slightly flexed position. This is based on the judgment that the user is standing on the top of a step ladder. (5) When the load on the knee joint 20 is released and backward acceleration is detected, the lock on the extension side is released, and only the lock on the flexion side is maintained. From there, the fully extended locked state is maintained until the mode release instruction is given. This is based on the judgment that the user is in the process of descending the stepladder. If, after (4), it is determined that the object is still rising, you may return to (2) and (3).

[0067] The determination unit 110 determines the suitability of the generated control command information output by the control command output unit 108 based on at least one of the operation request and constraint information 152. If the determination unit 110 determines that the generated control command information is unsuitable (Y in S26), the notification unit 112 notifies the user of the prosthetic leg 10 to that effect (S28), and the request acquisition unit 162 acquires an operation request again (S20). On the other hand, if the determination unit 110 determines that the generated control command information is not unsuitable (N in S26), the process proceeds to step S30.

[0068] The simulation unit 114 uses the generation control command information to simulate the control of the prosthetic leg and determines whether or not an error occurs. If no error occurs (N in S30), the process proceeds to step S34. If an error occurs (Y in S30), the correction unit 116 corrects the generation control command information. If the correction unit 116 cannot correct the generation control command information (N in S32), the notification unit 112 notifies the user of the prosthetic leg 10 of this fact (S28), and the request acquisition unit 162 acquires the operation request again (S20). On the other hand, if the correction unit 116 can correct the generation control command information (Y in S32), the process proceeds to step S34.

[0069] Server 2 transmits the corrected generation control command information, if corrected in step S32, or the generation control command information generated in step S24, to the control device 50 of the prosthetic leg 10 via the communication terminal 4. In the control device 50, the setting unit 182 causes the control unit 184 to set the generation control command information, and the control unit 184 controls the control mechanism 40 to control the operation of the prosthetic leg 10 according to the generation control command information (S34). In Server 2, the model update unit 118 updates the generation model 106 based on at least one of the simulation results of prosthetic leg control using the generation control command information by the simulation unit 114, and the actual control results of the prosthetic leg 10 using the generation control command information (S36).

[0070] As described above, the prosthetic leg control command generation system 1 of this embodiment includes an information acquisition unit 102 that acquires control-related information 150 concerning the operation control of an electronically controlled prosthetic leg 10, and a model generation unit 104 that generates a generation model 106 that takes an operation request of the prosthetic leg 10 as input and outputs control command information for the prosthetic leg 10, based on the information acquired by the information acquisition unit 102. As a result, when an operation request of the prosthetic leg 10 is input, a generation model 106 that outputs control command information for the prosthetic leg 10 can be generated, and control command information for controlling the prosthetic leg 10 in accordance with the requested operation can be quickly generated using the generated generation model 106. Therefore, the prosthetic leg 10 can be quickly controlled in accordance with the requested operation without having to update the firmware of the prosthetic leg 10 or the like.

[0071] In the prosthetic leg control command generation system 1, the control-related information 150 may be information that defines at least one of the range of achievable motion control and state of the prosthetic leg 10, the information acquisition unit 102 may further acquire constraint information 152 that restricts at least one of the range of motion control and state of the prosthetic leg 10, and the model generation unit 104 may further generate the generated model 106 based on the constraint information 152. Because the prosthetic leg 10 has constraints on motion and state, it was necessary to consider these constraints when generating control command information corresponding to the required motion. With this configuration, control command information can be generated with a high degree of freedom while taking these constraints into consideration.

[0072] The prosthetic leg control command generation system 1 may further include a request acquisition unit 162 that acquires operation requests for the prosthetic leg 10 from the user, and a control command output unit 108 that inputs the operation requests acquired by the request acquisition unit 162 to the generation model 106 and outputs the control command information output from the generation model 106 as generated control command information. This allows for the rapid output of control command information for the prosthetic leg 10 in response to operation requests from the user of the prosthetic leg 10.

[0073] In the prosthetic leg control command generation system 1, the request acquisition unit 162 may acquire input sentences in natural language as operation requests, and the generation model 106 may be capable of processing input sentences in natural language. This allows users to easily input operation requests using natural language.

[0074] The prosthetic leg control command generation system 1 may further include a setting unit 182 that causes the generated control command information to be set in the control unit 184 of the prosthetic leg 10. This allows the prosthetic leg 10 to be actually controlled according to the generated control command information.

[0075] The prosthetic leg control command generation system 1 may further include a simulation unit 114 that performs a simulation of prosthetic leg control using the generated control command information, and a modification unit 116 that modifies the generated control command information based on the simulation results from the simulation unit 114. This allows the generated control command information to be appropriately modified in response to program inconsistencies, etc.

[0076] The prosthetic leg control command generation system 1 may further include a model update unit 118 that updates the generated model 106 based on at least one of the simulation results of prosthetic leg control using the generated control command information and the actual control results of the prosthetic leg 10 using the generated control command information. This allows the generated model 106 to be optimized.

[0077] The prosthetic leg control command generation system 1 may further include a determination unit 110 that determines at least one of the following: that the generated control command information does not conform to the operation request, and that the generated control command information violates the constraint information 152; and a notification unit 112 that notifies the user according to the determination result by the determination unit 110. This allows the user to be informed that it is not possible to control the prosthetic leg 10 in response to the operation request.

[0078] In the prosthetic leg control command generation system 1, multiple generation models 106 may be provided, each of the multiple generation models 106 may be associated with a different user, the model update unit 118 may update each of the multiple generation models 106 individually, and the system may further include a sharing unit 120 that shares the generation control command information generated using one of the multiple generation models 106 with the users of the other generation models 106. This allows generation control command information to be generated using a generation model 106 optimized for each user, and the generated generation control command information to be shared among users.

[0079] The following describes modifications of this embodiment. In the prosthetic leg control command generation system 1, any one or more of the components shown in Figure 2, which are included in the server 2, communication terminal 4, and control device 50 of the prosthetic leg 10, may be provided in another device. Furthermore, the prosthetic leg control command generation system 1 does not necessarily have to include all of the server 2, communication terminal 4, and prosthetic leg 10; it may consist of only one of these devices, or any two of these devices. For example, each function of the prosthetic leg control command generation system 1 may be performed by the knee joint 20 alone.

[0080] The motion control of the prosthetic leg 10, controlled by the generated control command information, is not limited to controlling the rotational resistance of the knee joint 20, but may also be, for example, the motion control of the foot 12.

[0081] In the embodiment, it was explained that the communication terminal 4 uses the request acquisition unit 162 to acquire natural language text from the user as an action request, transmits the acquired natural language text to the server 2, and the server 2 interprets the action request by processing the natural language text using the control command output unit 108. However, the request acquisition unit 162 may also process the input natural language text and acquire the processed information as an action request. This allows the control command output unit 108 to output control command information based on the action request even if the generative model 106 used by the control command output unit 108 does not support natural language processing.

[0082] The model generation unit 104 was described as generating the generated model 106 based on control-related information 150 and constraint information 152. However, the model generation unit 104 may generate the generated model 106 based on control-related information 150, without relying on constraint information 152.

[0083] The embodiments of the present invention have been described in detail above. The embodiments described above are merely examples of how to implement the present invention. The contents of the embodiments do not limit the technical scope of the present invention, and many design changes, such as changes, additions, and deletions of components, are possible as long as they do not depart from the spirit of the invention as defined in the claims. In the embodiments described above, such design changes are described with notations such as "of the embodiments" or "in the embodiments," but design changes may also be permitted in contents without such notations.

[0084] Any combination of the embodiments and modifications described above is also useful as an embodiment of the present invention. The new embodiments resulting from these combinations possess the combined effects of the respective embodiments and modifications.

[0085] In the embodiments disclosed herein, those in which multiple functions are provided in a distributed manner may have some or all of those multiple functions integrated into a single unit, and conversely, those in which multiple functions are integrated may have some or all of those functions provided in a distributed manner. Whether the functions are integrated or distributed, the configuration should be such that the objective of the invention can be achieved. [Explanation of symbols]

[0086] 1...Prosthetic leg control command generation system, 2...Server, 4...Communication terminal, 10...Prosthetic leg, 20...Knee joint, 102...Information acquisition unit, 104...Model generation unit, 106...Generated model, 108...Control command output unit, 110...Determination unit, 112...Notification unit, 114...Simulation unit, 116...Modification unit, 118...Model update unit, 120...Sharing unit, 150...Control-related information, 152...Constraint information, 162...Request acquisition unit, 182...Setting unit, 184...Control unit.

Claims

1. An information acquisition unit that acquires control-related information regarding the motion control of an electronically controlled prosthetic leg, A model generation unit generates a generation model that takes the motion request of the prosthetic leg as input and outputs control command information of the prosthetic leg, based on the information acquired by the information acquisition unit. A model generation system for prosthetic limb control that includes the following features.

2. The control-related information is information that defines at least one of the range of achievable motion controls and states of the prosthetic leg, The information acquisition unit further acquires constraint information that limits at least one of the range of motion control and state of the prosthetic leg. The model generation unit generates the generated model based on the constraint information. The prosthetic leg control model generation system according to claim 1.

3. A request acquisition unit that acquires operation requests for the prosthetic leg from the user, A control command output unit inputs the operation request acquired by the request acquisition unit to the generation model, and outputs the control command information output from the generation model as generated control command information. The prosthetic leg control model generation system according to claim 1 or 2, further comprising:

4. The request acquisition unit acquires the input sentence in natural language as the operation request, The generative model is capable of natural language processing of the input sentence. The prosthetic leg control model generation system according to claim 3.

5. The prosthetic leg control model generation system according to claim 3, further comprising a setting unit for setting the generation control command information to the control unit of the prosthetic leg.

6. A simulation unit that performs a simulation of prosthetic leg control using the aforementioned generation and control command information, A modification unit modifies the generation control command information based on the simulation results from the simulation unit, The prosthetic leg control model generation system according to claim 3, further comprising:

7. The model update unit further includes a model update unit that updates the generation model based on at least one of the simulation results of prosthetic leg control using the generation control command information and the actual control results of the prosthetic leg using the generation control command information. The prosthetic leg control model generation system according to claim 3.

8. A determination unit that determines at least one of the following: that the generation control command information does not conform to the operation request, and that the generation control command information violates constraint information that restricts at least one of the range of operation control and state of the prosthetic leg; A notification unit that notifies the user according to the determination result by the determination unit, The prosthetic leg control model generation system according to claim 3, further comprising the above.

9. Multiple generation models are provided, and each of these generation models is associated with a different user. The model update unit updates each of the plurality of generation models individually. The system further includes a sharing unit that shares generation control command information generated using one of the multiple generation models with users of other generation models. The prosthetic leg control model generation system according to claim 7.

10. The system comprises a server, a communication terminal capable of communicating with the server, and a prosthetic leg capable of communicating with the communication terminal. The server comprises an information acquisition unit that acquires control-related information concerning the motion control of an electronically controlled prosthetic leg, and a model generation unit that generates a generation model that takes the motion request of the prosthetic leg as input and outputs control command information of the prosthetic leg, based on the information acquired by the information acquisition unit. The communication terminal includes a request acquisition unit that acquires operation requests for the prosthetic leg from the user. The server further comprises a control command output unit that inputs the operation request acquired by the request acquisition unit to the generation model and outputs the control command information output from the generation model as generated control command information. The aforementioned prosthetic leg comprises a control unit that controls the operation of the device in accordance with control command information, and a setting unit that causes the control unit to set the generated control command information. Prosthetic leg control command generation system.

11. An electronically controlled knee joint comprising a control unit that controls rotational resistance around the knee axis according to control command information, An information acquisition unit that acquires control-related information concerning the operation control of the knee joint, A model generation unit generates a generation model that takes the operation request of the knee joint as input and outputs control command information of the knee joint, based on the information acquired by the information acquisition unit. A request acquisition unit that acquires an operation request for the knee joint from the user, A control command output unit inputs the operation request acquired by the request acquisition unit to the generation model, and outputs the control command information output from the generation model as generated control command information. A setting unit that causes the control unit to set the generation control command information, A knee joint equipped with a knee joint.

12. A method for generating a model for prosthetic leg control, which is performed by a computer, An acquisition step to obtain control-related information regarding the motion control of an electronically controlled prosthetic leg, A generation step that generates a generation model that takes the motion request of the prosthetic leg as input and outputs control command information of the prosthetic leg, based on the information acquired in the acquisition step, A method for generating a model for controlling a prosthetic leg, including the above.

13. An acquisition step to obtain control-related information regarding the motion control of an electronically controlled prosthetic leg, A generation step that generates a generation model that takes the motion request of the prosthetic leg as input and outputs control command information of the prosthetic leg, based on the information acquired in the acquisition step, A program for generating models for prosthetic leg control, which is used to run on a computer.

Images