Method, device, terminal equipment and storage medium for determining hip joint assistance value

Abstract

The embodiments of the present application are applicable to the field of computer technology, and provide a method, apparatus, terminal device, and storage medium for determining the power assist value of a hip joint, the method comprising: in response to the power assist calculation task of either side of the hip joint in the current state, calculating the user's first angle data based on a walking trajectory function and a preset time offset constant; the walking trajectory function is constructed by at least two historical angle data that have been collected; and determining the power assist value of either side of the hip joint in the current state based on the first angle data and the second angle data of either side of the hip joint in the current state. Through the method provided in this embodiment, the terminal device can continuously update the power assist value of the hip joint exoskeleton according to the user's walking trajectory function and the first angle data at a certain moment in the future, so that the hip joint exoskeleton can provide the user with appropriate and variable assist force according to the user's walking conditions.

Description

Technical Field

[0001] The embodiments of the present application belong to the field of computer technology, and in particular, relate to a method, apparatus, terminal device and storage medium for determining a hip joint assistance value. Background Art

[0002] The lower limb exoskeleton robot is mainly composed of a sensor module, a power output module, a control system and an energy system. The lower limb exoskeleton robot can obtain the motion data of the user's hip joint through the sensor module, and the control system in the lower limb exoskeleton robot determines whether the user is in a walking state based on the acquired motion data. When the control system in the lower limb exoskeleton robot determines that the user is in a walking state, the power output module can be controlled to apply auxiliary force to the left and right sides of the user's hip joint to assist the user in walking.

[0003] However, in the prior art, the assisting force provided by the lower limb exoskeleton robot to the user is constant and is only applicable to the user walking at a constant speed. When the user changes the speed of walking, the frequency and amplitude of the hip joint torque will continue to change, so the optimal assisting force required by the hip joint will also change accordingly. Therefore, the existing assisting method of the lower limb exoskeleton robot cannot provide the user with variable assisting force when the user changes the speed of walking. Summary of the invention

[0004] In view of this, the embodiments of the present application provide a method, apparatus, terminal device and storage medium for determining a hip joint assistance value, so as to improve the provision of appropriate and variable assisting force to the user according to the user's walking conditions.

[0005] A first aspect of an embodiment of the present application provides a method for determining a hip joint assistance value, comprising:

[0006] In response to the power calculation task on either side of the hip joint in the current state, first angle data of the user is calculated based on a walking trajectory function and a preset time offset constant; the walking trajectory function is constructed by at least two collected historical angle data;

[0007] The assist value of either side of the hip joint in the current state is determined based on the first angle data and the second angle data of either side of the hip joint in the current state.

[0008] In a possible implementation of the first aspect, determining the assist value of either side of the hip joint in the current state based on the first angle data and the second angle data of either side of the hip joint in the current state includes:

[0009] determining an angle difference based on the first angle data and the second angle data;

[0010] Inputting the angle difference into a torque estimation model to determine a torque estimation value;

[0011] The assist value is determined based on a preset gain coefficient and the torque estimation value.

[0012] In a possible implementation manner of the first aspect, before inputting the angle difference into a torque estimation model to determine a torque estimation value, the method further includes:

[0013] Obtaining a model to be trained and a training sample input by a user, and processing the training sample through the model to be trained to generate an initial torque value; the training sample includes a plurality of training angle values ​​and an expected torque value corresponding to each training angle value;

[0014] Determine a loss value between the initial torque value and the expected torque value based on a preset first loss function;

[0015] The model to be trained is updated based on the loss value until the loss value satisfies a preset training stop condition, and the model to be trained corresponding to when the loss value satisfies the training stop condition is used as the torque estimation model.

[0016] In a possible implementation manner of the first aspect, after updating the model to be trained based on the loss value until the loss value satisfies a preset training stop condition, and using the model to be trained corresponding to when the loss value satisfies the training stop condition as the moment estimation model, the method further includes:

[0017] Inputting the torque estimation model and the training sample into a preset evaluation function to generate a torque accuracy function; the torque accuracy function is used to determine the corresponding relationship between torque and duration;

[0018] Solving the minimum torque for the torque accurate function to determine the minimum torque accurate value corresponding to the torque accurate function;

[0019] The duration corresponding to the minimum torque accurate value is used as the time offset constant.

[0020] In a possible implementation of the first aspect, before the calculation of the first angle data of the user based on the walking trajectory function and a preset time offset constant in response to the power assistance calculation task on either side of the hip joint in the current state, the method includes:

[0021] Determine maximum angle data and minimum angle data from the plurality of historical angle data;

[0022] determining an amplitude coefficient based on the maximum angle data and the minimum angle data;

[0023] Inputting the plurality of historical angle data into a preset second loss function to calculate a phase coefficient;

[0024] Inputting the plurality of historical angle data into a preset frequency function to calculate a frequency variation coefficient;

[0025] The walking trajectory function is established based on the amplitude coefficient, the phase coefficient and the frequency change coefficient.

[0026] In a possible implementation manner of the first aspect, before determining the maximum angle data and the minimum angle data from the plurality of historical angle data, the method includes:

[0027] Acquire multiple historical walking angles of any side of the hip joint within a preset statistical time period before the corresponding moment of the current state;

[0028] Inputting the multiple historical walking angles into a preset normalization algorithm to generate normalized data corresponding to each of the historical walking angles;

[0029] The historical angle data of either side of the hip joint in N gait cycles is determined from the normalized data; wherein N is a positive integer greater than or equal to 2.

[0030] In a possible implementation manner of the first aspect, determining the plurality of historical angle data on either side of the hip joint in N gait cycles from the normalized data includes:

[0031] Inputting the plurality of normalized data and the acquisition time corresponding to each of the normalized data into a preset cycle algorithm to determine the duration of the gait cycle;

[0032] The historical angle data of either side of the hip joint in the N gait cycles are determined from the normalized data based on the gait cycle duration.

[0033] A second aspect of an embodiment of the present application provides a device for determining a hip joint assistance value, comprising:

[0034] An angle data calculation module, for calculating the first angle data of the user based on a walking trajectory function and a preset time offset constant in response to the power assistance calculation task on either side of the hip joint in the current state; the walking trajectory function is constructed by at least two collected historical angle data;

[0035] The assist value determination module is used to determine the assist value of either side of the hip joint in the current state based on the first angle data and the second angle data of either side of the hip joint in the current state.

[0036] A third aspect of an embodiment of the present application provides a terminal device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for determining the hip joint assistance value as described in the first aspect above is implemented.

[0037] A fourth aspect of an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for determining the hip joint assistance value as described in the first aspect above is implemented.

[0038] A fifth aspect of the embodiments of the present application provides a computer program product. When the computer program product is run on a computer, the computer executes the method for determining the hip joint assistance value described in the first aspect.

[0039] Compared with the prior art, the embodiments of the present application have the following advantages:

[0040] In the embodiment of the present application, the terminal device can respond to the power calculation task on either side of the user's hip joint in the current state, and calculate the user's first angle data based on the walking trajectory function of the hip joint on that side and the time offset constant pre-set by the developer; wherein the walking trajectory function is constructed by the terminal device through at least two historical angle data of the hip joint that have been collected; after calculating the first angle data, the terminal device can determine the power value of the hip joint in the current state according to the first angle data and the second angle data of the hip joint on that side in the current state. In this embodiment, the terminal device can calculate the first angle data of the hip joint on that side at a certain moment in the future according to the walking trajectory function and the time offset constant of one side of the user's hip joint, and determine the power value according to the first angle data at a certain moment in the future and the second angle data in the current state. The method provided in this embodiment enables the hip joint exoskeleton to provide the user with appropriate and variable auxiliary force according to the user's walking situation, realizes the real-time variability of the power value, and improves the flexibility of the power; and predicts the power value of the future state through the time offset constant, thereby improving the accuracy of the power assistance of the hip joint exoskeleton. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments or prior art descriptions. Obviously, the drawings described below are only some embodiments of the present application, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0042] Figure 1 is a schematic diagram of a method for determining a hip joint assistance value provided in an embodiment of the present application;

[0043] Figure 2 It is a schematic diagram of angle data provided in an embodiment of the present application;

[0044] Figure 3 This is a schematic diagram of the power assistance process of a lower limb exoskeleton robot provided in an embodiment of the present application;

[0045] Figure 4 It is a schematic diagram of a walking trajectory curve and a power assistance curve provided in an embodiment of the present application;

[0046] Figure 5 is a schematic diagram of another method for determining a hip joint assistance value provided in an embodiment of the present application;

[0047] Figure 6 is a schematic diagram of a device for determining a hip joint assistance value provided in an embodiment of the present application;

[0048] Figure 7 It is a schematic diagram of a terminal device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0049] In the following description, specific details such as specific system structures, technologies, etc. are proposed for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from hindering the description of the present application.

[0050] The walking speed of people in daily life varies within a wide range of speeds. When the walking speed is low, the human hip joint has a smaller torque and a lower frequency of change; when the walking speed is high, the human hip joint has a higher torque and a higher frequency of change. In order to assist people in walking in daily environments, a series of lower limb exoskeleton robot prototypes have been developed in recent years. They can provide people with a relatively stable walking assist force at a constant speed, so as to achieve the effect of reducing human metabolism to a certain extent. However, when people walk at variable speeds, the frequency and amplitude of the walking joint torque change, and the optimal assist force required also changes accordingly, resulting in the existing lower limb exoskeleton robots that provide fixed assist force cannot adapt well to the assist situation under walking speed, which makes the development of an assist exoskeleton that can quickly adapt to variable speed walking very challenging.

[0051] In the current research on lower limb exoskeleton robot control strategies, the methods of determining the power assist value based on the power assist curve can be mainly divided into two categories: the first category is the control method based on the preset curve; the second category is the control method based on the real-time mapping curve.

[0052] In the control method based on the preset curve, the terminal device usually needs to calculate the power-assistance curve in the next gait cycle in advance, and trigger the corresponding power-assistance mode according to the discrete gait event detection. Among them, the power-assistance curve calculated by the terminal device can be a curve reflecting the relationship between time and the torque of the hip joint, or a curve reflecting the relationship between time and the position of the hip joint. However, triggering the power-assistance mode through discrete gait events will inevitably lead to the failure of the power-assistance mode to be triggered under certain special circumstances, thereby causing the failure of the power-assistance in a certain gait cycle. Secondly, by triggering the power-assistance through gait events, it is difficult to ensure that the auxiliary force of the preset curve is consistent with the walking gait. Once the auxiliary force provided by the lower limb exoskeleton robot conflicts with walking, it is very easy for the user to become unstable or even fall.

[0053] In order to solve this problem, Medrano et al. introduced continuous phase estimation into the research of exoskeleton robot control strategy. After the continuous phase is introduced, the terminal device can estimate the gait phase of a person when walking through the inertial measurement of the hip joint, thereby changing the gait event of the hip joint from discrete to continuous. Finally, by estimating the walking phase in real time, the terminal device can find out the size of the power assist corresponding to the current phase. After the continuous phase is introduced, the problem of failure to trigger the power assist mode is solved to a certain extent. However, the control method based on the preset curve is limited by the preset curve. It is not only difficult to adapt to the differences in the form of power assist required by different users, but also difficult to cope with gait changes during daily walking (such as starting, emergency stop, and speed change).

[0054] In the control method based on real-time mapping curve, the terminal device can continuously obtain the power torque required at the current moment by real-time mapping of a continuous signal during walking. Therefore, this method has good adaptability between different users and gait changes. Gasparri et al. established a mapping model for real-time ankle torque estimation, which can use ground reaction force to approximate the net muscle torque and calculate the required ankle power torque in real time based on the ground reaction force. Bishe et al. further studied this, estimated the ankle torque through inverse kinematic torque balance, and scaled the ankle torque to determine the power value. This method can more accurately estimate the joint torque during walking in real time. Since this method needs to calculate the hip torque through the ankle torque, however, during the human walking process, the ankle joint has two actions: swinging and supporting, and the hip joint power is only related to the swing of the leg. Therefore, this method needs to use a finite state machine to distinguish the support phase and swing phase of the ankle joint. Therefore, the accuracy of the power value calculated by the ankle torque is limited by the accuracy of the developer's classification. In this regard, Tan et al. established a time-independent ankle torque mapping model, which can calculate the ankle torque in real time through the calf angle, thereby realizing continuous mapping of joint torque. However, the time-independent ankle torque mapping model requires adjusting the mapping model parameters through pressure insoles, and it is difficult to effectively adapt to sudden non-rhythmic gaits such as sudden stops and speed changes.

[0055] In order to further improve the control strategy of the lower limb exoskeleton robot, Lim et al. reduced the sensors of the lower limb exoskeleton robot to the hip joint encoders on both sides. The encoders of the hip joints on both sides can map the hip joint assist torque required by the current user through the angle difference between the hip joints on both sides. This method can effectively help users walk in daily life and can better adapt to non-rhythmic gaits. However, although the mapped torque can adapt well to the walking frequency at different walking speeds, it fails to adjust the magnitude of the assist value under the variable speed walking state. When people walk at variable speeds, the optimal assist value they need often changes. Therefore, how to use a small number of sensors to accurately estimate the hip joint torque and hip joint assist value under variable speed walking state, and resist the inevitable interference caused by non-rhythmic gait and exoskeleton hardware to achieve robust control in variable speed walking scenarios is still a very challenging topic.

[0056] The technical solution of the present application is described below through specific embodiments.

[0057] Reference Figure 1, showing a schematic diagram of a method for determining a hip joint assist value provided by an embodiment of the present application. The method for determining a hip joint assist value can be applied to a control terminal of a lower limb exoskeleton robot, which can be a single-chip microcomputer, a smart phone, a computer, a programmable logic controller, or other terminal device. The above method can specifically include the following steps:

[0058] S101. In response to the power assistance calculation task on either side of the hip joint in the current state, first angle data of the user is calculated based on a walking trajectory function and a preset time offset constant; the walking trajectory function is constructed by at least two collected historical angle data.

[0059] In this embodiment, when the user needs the lower limb exoskeleton robot to provide auxiliary force, the user can start the lower limb exoskeleton robot by pressing the start button on the lower limb exoskeleton robot. After the lower limb exoskeleton robot is started, the encoders fixed on both sides of the user's waist can be used to obtain the historical angle data of the left side of the user's hip joint and the historical angle data of the right side of the user's hip joint, and initiate a power calculation task to the terminal device to determine the power value corresponding to each hip joint. Figure 2 , shows a schematic diagram of angle data provided by an embodiment of the present application. Figure 2 As shown, an encoder a may be fixed on the right side of the user's waist. The encoder a on the right side of the waist may be connected to a locator a on the user's right thigh to obtain angle data θ of the user's right hip joint.

[0060] The terminal device can respond to the power assistance calculation task in the current state, and perform power assistance calculation for both sides of the hip joint according to the historical angle data of the left side of the hip joint and the historical angle data of the right side of the hip joint, so as to respectively determine the power assistance value of the left side of the user's hip joint and the power assistance value of the right side of the user's hip joint. For any side of the user's hip joint, the terminal device can construct a walking trajectory function of the hip joint on this side according to at least two historical angle data collected corresponding to this side. After constructing the walking trajectory function, the terminal device can calculate the first angle data of the hip joint on this side according to the time offset constant pre-set by the developer. Among them, the first angle data calculated by the terminal device can be used to represent the angle of the hip joint on this side at a certain moment in the future corresponding to the time offset constant.

[0061] S102, determining the power assist value of either side of the hip joint in the current state based on the first angle data and the second angle data of either side of the hip joint in the current state.

[0062] In this embodiment, after calculating the first angle data of the user's hip joint on one side, the terminal device can obtain the second angle data of the hip joint on the side in the current state through the encoder on the hip joint on the side. The terminal device can determine the power assist value corresponding to the hip joint on the side based on the first angle data and the second angle data of the hip joint on the side in the current state.

[0063] In this embodiment, the terminal device can directly generate the first angle data of the user at a future moment according to the walking trajectory function and the time bias constant. Therefore, through the method provided in this embodiment, even in a variable speed walking scenario, the terminal device can still accurately estimate the walking trajectory of the hip joint. In addition, since the assist value is generated by the terminal device based on the second angle data of the user at the current moment and the first angle data at the future moment. Therefore, through the method provided in the embodiment of the present application, the lower limb exoskeleton robot can resist non-rhythmic gait and the inevitable interference caused by the exoskeleton hardware to achieve robust control in a variable speed walking scenario.

[0064] In a possible implementation, after calculating the first angle data, the terminal device can determine the angle difference between the current state of the hip joint on this side and a certain moment in the future according to the first angle data and the second angle data. After calculating the angle difference, the terminal device can input the calculated angle difference into the torque estimation model pre-set by the developer to generate the torque estimation value of the hip joint on this side through the torque estimation model. Among them, the torque estimation value generated by the terminal device can be used to represent the auxiliary torque required by the hip joint on this side. After calculating the torque estimation value, the terminal device can generate the power assist value corresponding to the hip joint on this side according to the gain coefficient and the torque estimation value pre-set by the user. Among them, the gain coefficient can be adjusted by the user according to the usage habits. The gain coefficient can be used to scale the torque estimation value so that the auxiliary force output by the lower limb exoskeleton robot is more in line with the user's usage habits. After determining the power assist value of the hip joint on this side, the terminal device can transmit the power assist value corresponding to the hip joint on this side to the power output module of the lower limb exoskeleton robot, so as to provide the user with the power assist value The corresponding auxiliary force of the hip joint on this side is provided through the power output module.

[0065] See also Figure 3 , shows a schematic diagram of the assistance process of a lower limb exoskeleton robot provided in an embodiment of the present application. Figure 3As shown, when the user starts walking, the lower limb exoskeleton robot can obtain the user's historical angle data through the data acquisition module. The data acquisition module can input the acquired historical angle data into the state prediction module of the lower limb exoskeleton robot to generate the user's first angle data through the state prediction module. Then, the state prediction module can input the generated first angle data into the torque estimation module, and the torque estimation module can generate a torque estimation value based on the user's second angle data at the current moment and the first angle data at a certain moment in the future. The torque estimation module can input the generated torque estimation value into the power assist value calculation module. The power assist value calculation module can generate a power assist value based on the gain parameter and the torque estimation value set by the user. Finally, the lower limb exoskeleton robot can apply auxiliary force to both sides of the user's hip joint according to the power assist values ​​corresponding to both sides of the hip joint, so as to achieve assisted walking for the user.

[0066] like Figure 4 The figure shows a walking trajectory curve and a power assistance curve provided by an embodiment of the present application. Figure 4 (a) in the figure can be a walking trajectory curve of a walking trajectory function established by a terminal device based on multiple historical angle data. The solid line portion of the walking trajectory curve can be composed of multiple collected historical angle data, which can represent the walking trajectory of the user. The dotted line portion of the walking trajectory curve can include the first angle data at a certain moment in the future, which can represent the walking trajectory of the user at a future moment predicted by the terminal device based on the walking trajectory function and the time offset constant. Figure 4 (b) in the figure may be a power assist curve of a lower limb exoskeleton robot. The darker power assist curve may be composed of multiple torque estimation values ​​calculated by the terminal device. The lighter power assist curve may be composed of the power assist value generated by the terminal device after scaling the torque estimation value by the gain coefficient.

[0067] In this embodiment, after calculating the torque estimation value, the terminal device can determine the final assistance value according to the gain coefficient set by the user. Therefore, the method provided in this embodiment can ensure that the assistance force provided by the lower limb exoskeleton robot meets the user's usage habits, so as to provide the user with appropriate walking assistance force.

[0068] In a possible implementation, the calculation formula of the angle difference may be as follows.

[0069] Δθ(Δt)=θ(t+Δt)-θ t

[0070] Wherein, Δθ(Δt) may represent the angle difference between the first angle data and the second angle data. Δt may represent the time offset constant. θ(t+Δt) may represent the first angle data calculated according to the walking trajectory function and the time offset constant, where θ(t) may represent the walking trajectory function, and Δt may represent the time offset constant. θ t It can represent the second angle data of the hip joint on this side in the current state.

[0071] In a possible implementation, a formula for calculating the assist value according to the gain coefficient and the torque estimation value may be as follows.

[0072] τ=kM(Δθ)

[0073] k∈[0,0.1]

[0074] Wherein, τ may represent the assist value. M(Δθ) may represent the torque estimation value output by the torque estimation model. k may represent the gain coefficient set by the user, and the value range of the gain coefficient may be 0 to 0.1.

[0075] In a possible implementation, the function expression of the torque estimation model can be as follows.

[0076] M(Δθ)=(Δθ) 2 sign(Δθ)

[0077]

[0078] Wherein, Δθ may be the angle difference between the first angle data and the second angle data. M(Δθ) may be the torque estimation value corresponding to the angle difference. * It can be the first weighted constant coefficient. * It can be the second weighted constant coefficient. The first weighted constant coefficient and the second weighted constant coefficient can be trained by developers through training samples during research and development.

[0079] In a possible implementation, before a user uses a lower limb exoskeleton robot, a developer may first train a model to be trained using training samples to generate a torque estimation model. The developer may input the training samples and the model to be trained into a terminal device. The terminal device may obtain the model to be trained and the training samples input by the developer, and process the training samples using the model to be trained to generate an initial torque value. The training samples may include multiple training angle values ​​and the expected torque values ​​corresponding to each training angle value.

[0080] After calculating the initial torque values ​​corresponding to each training angle value, the terminal device can determine the loss value between the initial torque value and the expected torque value according to the first loss function preset by the developer. After calculating the loss value, the terminal device can update the model to be trained according to the loss value until the loss value meets the training stop condition preset by the developer. At this time, the terminal device can use the model to be trained when the loss value meets the training stop condition as the torque estimation model.

[0081] In this embodiment, since the torque estimation model is trained by training samples according to the training stop condition, the torque estimation model generated by this embodiment can accurately generate the torque estimation value.

[0082] In a possible implementation, the first loss function may be as follows.

[0083]

[0084] Among them, Acc(τ est ) can represent the loss value between the initial torque value and the expected torque value. Acc can represent the root mean square error (RMSE). n can represent the number of training angle values ​​in the training sample. τ esti It can represent the initial torque value corresponding to the i-th training angle value. τ truei It can represent the expected torque value corresponding to the i-th training angle value.

[0085] In a possible implementation, the terminal device can use the model to be trained when the loss value is minimized as the moment estimation model, and use the values ​​of the two weighted constant coefficients in the model at this time as the optimal values ​​of the first weighted constant coefficient and the second weighted constant coefficient. At this time, the function expression of the training stop condition can be as follows.

[0086]

[0087] Among them, a * It can represent the first weighted constant coefficient in the moment estimation model. * It can represent the second weighted constant coefficient in the moment estimation model. Acc(τ est ) can represent the loss value between the initial torque value and the expected torque value.

[0088] In a possible implementation, after generating the torque estimation model, the terminal device may also generate a time bias constant based on the torque estimation model and the training sample. The terminal device may input the generated torque estimation model and the training sample into the evaluation function to generate a torque accuracy function. Among them, the torque accuracy function can be used to determine the corresponding relationship between the torque and the duration. Then, the terminal device may perform a minimum torque solution on the torque accuracy function to determine the minimum torque accuracy value corresponding to the torque accuracy function. The terminal device may use the duration corresponding to the minimum torque accuracy value as a time bias constant to calculate the first angle data of the user at a certain moment in the future.

[0089] In a possible implementation, the calculation formula of the time offset constant may be as follows.

[0090]

[0091] Here, Δt may represent the calculated time offset constant. It can represent the minimum moment solution of the moment accuracy function. G(Δθ(x)) can represent the moment accuracy function. Specifically, the evaluation function in the terminal device can be any function known to those skilled in the art that can be used to evaluate effectiveness, such as a root mean square error function, and this embodiment is not intended to be specifically limited to this.

[0092] Figure 5 FIG. 1 shows a specific implementation flow chart of a method S101 for determining a hip joint assistance value provided in the second embodiment of the present application. Figure 5 , compared to Figure 1 In the embodiment, the method for determining the hip joint assistance value provided in the embodiment further includes, before S101, S501 to S505, which are described in detail as follows:

[0093] S501. Determine maximum angle data and minimum angle data from a plurality of historical angle data.

[0094] In this embodiment, before calculating the first angle data through the walking trajectory function and the time offset constant, the terminal device can first construct the walking trajectory function of the hip joint on the side according to at least two historical angle data of the hip joint on the side. The terminal device can determine the maximum angle data and the minimum angle data from the multiple historical angle data of the hip joint on the side.

[0095] In one possible implementation, after the user starts the lower limb exoskeleton robot, the lower limb exoskeleton robot can collect the historical walking angles of the user's hip joints on both sides through the encoders on both sides of the user's waist, and continuously send the collected historical walking angles to the terminal device to generate a walking trajectory function on each side of the hip joint. For any hip joint on one side of the user, the terminal device can obtain multiple historical walking angles that have been collected according to the statistical time preset by the developer. After obtaining multiple historical walking angles, the terminal device can input the multiple historical walking angles obtained into the normalization algorithm preset by the developer to generate normalized data corresponding to each historical walking angle.

[0096] In one possible implementation, the normalization algorithm may be as follows.

[0097]

[0098] S norm =S old -c

[0099] Among them, maxS old It can represent the maximum value of multiple historical walking angles corresponding to the statistical duration. minS old It can represent the minimum value of multiple historical walking angles corresponding to the statistical duration. c can represent the normalization coefficient. S normi and S oldi are all one-dimensional vectors. norm It can represent a data set consisting of multiple normalized data corresponding to the statistical duration, that is, a normalized data set. old It can represent a data set consisting of multiple historical walking angles corresponding to the statistical duration, that is, a walking data set.

[0100] After generating normalized data corresponding to each historical walking angle, the terminal device can input multiple normalized data and the collection time corresponding to each normalized data into the cycle algorithm pre-set by the developer to determine the gait cycle duration corresponding to the hip joint on that side through the cycle algorithm.

[0101] In a possible implementation, the periodic algorithm may be specifically as follows.

[0102] T=abs(t zero1 -t zero2 )·2

[0103] Wherein, T can represent the duration of the gait cycle. abs can represent the absolute value operator. t zero1 It can represent the time value corresponding to the normalized data with an angle value of 0 and the latest acquisition time in the normalized data set. zero2It can represent the time value corresponding to the normalized data with an angle value of 0 and the second latest acquisition time in the normalized data set. When there are no two normalized data with an angle value of 0 in the normalized data set, the gait cycle duration T can be 1.5 seconds.

[0104] In this embodiment, since the terminal device can calculate the user's gait cycle duration in real time based on multiple historical angle data, through the method provided by this embodiment, even when the user's gait periodicity is irregular, the lower limb exoskeleton robot can stably provide auxiliary force similar to the human walking trajectory.

[0105] After calculating the gait cycle duration corresponding to the hip joint on that side, the terminal device can determine the historical angle data of the hip joint on that side in N gait cycles from the normalized data according to the number of gait cycles N and the gait cycle duration preset by the developer. Among them, any hip joint on one side of the user can contain multiple historical angle data in one gait cycle.

[0106] In a possible implementation, a data set of historical angle data of a hip joint on one side in N gait cycles may be as follows.

[0107] S′=[S norm (t-NT),S norm (t-NT+1),...,S norm (t)]

[0108] Among them, S norm (t) can represent the normalized data at the current moment. norm (t-NT) may represent normalized data before the duration of N gait cycles. S′ may be a one-dimensional vector, which may be used to represent a data set consisting of all historical angle data in N gait cycles, namely, a historical data set.

[0109] S502: Determine an amplitude coefficient based on the maximum angle data and the minimum angle data.

[0110] In this embodiment, after determining the maximum angle data and the minimum angle data, the terminal device may input the maximum angle data and the minimum angle data into an amplitude calculation formula preset by a developer to determine the amplitude coefficient in the walking trajectory function.

[0111] In a possible implementation, the amplitude calculation formula may be as follows.

[0112]

[0113] Wherein, A may represent the amplitude coefficient, maxS′ may represent the maximum angle data in the historical data set, and minS′ may represent the minimum angle data in the historical data set.

[0114] S503: Input the plurality of historical angle data into a preset second loss function to calculate a phase coefficient.

[0115] In this embodiment, the terminal device may also input multiple historical angle data into the second loss function preset by the developer to calculate the phase coefficient corresponding to the hip joint on that side. Specifically, the terminal device may input multiple historical angle data into the second loss function to generate a phase accuracy function, and then the terminal device may perform a minimum loss solution on the phase accuracy function to determine the minimum phase loss value corresponding to the phase accuracy function. The terminal device may use the phase corresponding to the minimum phase loss value as the phase coefficient.

[0116] In one possible implementation, the second loss function may be as follows.

[0117]

[0118] Wherein, λ may represent a phase loss coefficient, and by way of example, the phase loss coefficient λ may be 10. A may represent an amplitude coefficient. It can represent the frequency variation coefficient. may represent a phase. c may represent a normalization coefficient. S(k) may represent the kth historical angle data in the historical angle data set corresponding to the most recent gait cycle. T may represent the duration of the gait cycle. exp may represent an exponential function.

[0119] In a possible implementation manner, the phase coefficient may be calculated using the following formula.

[0120]

[0121] in, Can represent the phase coefficient. Loss can represent the second loss function. It can be expressed as solving the minimum loss for the second loss function.

[0122] S504: Input the plurality of historical angle data into a preset frequency function to calculate a frequency variation coefficient.

[0123] In this embodiment, the terminal device may also input a plurality of historical angle data into a frequency function preset by a developer to calculate a frequency change coefficient corresponding to the hip joint on that side.

[0124] In a possible implementation, a calculation formula for the frequency variation coefficient may be as follows.

[0125]

[0126] in, can represent the frequency variation coefficient. fft can represent the Fourier transform function. S′ can represent a data set consisting of all historical angle data in N gait cycles. It can be expressed as the maximum value of the Fourier transform solution for all historical angle data.

[0127] S505: Establish the walking trajectory function based on the amplitude coefficient, the phase coefficient and the frequency variation coefficient.

[0128] In this embodiment, after the terminal device calculates the amplitude coefficient, phase coefficient and frequency change coefficient respectively, it can establish the walking trajectory function corresponding to any side of the hip joint according to the amplitude coefficient, phase coefficient, frequency change coefficient and normalization coefficient of that side.

[0129] In a possible implementation, the walking trajectory function constructed by the terminal device may be as follows.

[0130]

[0131] Among them, A can represent the amplitude coefficient corresponding to a certain side of the hip joint. It can represent the frequency change coefficient corresponding to the hip joint on that side. It can represent the phase coefficient corresponding to the hip joint on this side. c can represent the normalization coefficient corresponding to the hip joint on this side.

[0132] In this embodiment, the terminal device can construct a walking trajectory function through the amplitude coefficient, phase coefficient and frequency change coefficient, and predict the first angle data of the user at a certain moment in the future through the walking trajectory function. Therefore, through the method provided by this embodiment, the lower limb exoskeleton robot does not need to estimate the gait phase during walking, nor does it need to identify the gait events of walking, which can significantly improve the adaptability of the lower limb exoskeleton robot to emergency stops, starts, and continuous walking speed changes. Furthermore, since the walking trajectory function is established based on the historical angle data of the user in N gait cycles, the walking trajectory function generated by this embodiment can accurately express the walking trajectory of the user in the past period of time, thereby improving the prediction accuracy of the future walking trajectory of the user of the lower limb exoskeleton robot.

[0133] It should be noted that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0134] Reference Figure 6, shows a schematic diagram of a device for determining a hip joint assist value provided by an embodiment of the present application, which may specifically include an angle data calculation module 601 and an assist value determination module 602, wherein:

[0135] Angle data calculation module 601, for calculating the first angle data of the user based on a walking trajectory function and a preset time offset constant in response to the power calculation task of either side of the hip joint in the current state; the walking trajectory function is constructed by at least two collected historical angle data;

[0136] The assist value determination module 602 is used to determine the assist value of either side of the hip joint in the current state based on the first angle data and the second angle data of either side of the hip joint in the current state.

[0137] The assist value determination module can also be used to determine the angle difference based on the first angle data and the second angle data; input the angle difference into the torque estimation model to determine the torque estimation value; and determine the assist value based on a preset gain coefficient and the torque estimation value.

[0138] The assist value determination module can also be used to obtain the model to be trained and the training sample input by the user, and process the training sample through the model to be trained to generate an initial torque value; the training sample includes multiple training angle values ​​and the expected torque value corresponding to each training angle value; based on a preset first loss function, the loss value between the initial torque value and the expected torque value is determined; based on the loss value, the model to be trained is updated until the loss value meets the preset training stop condition, and the model to be trained corresponding to the time when the loss value meets the training stop condition is used as the torque estimation model.

[0139] The assist value determination module can also be used to input the torque estimation model and the training sample into a preset evaluation function to generate a torque accuracy function; the torque accuracy function is used to determine the corresponding relationship between torque and duration; the minimum torque is solved for the torque accuracy function to determine the minimum torque accuracy value corresponding to the torque accuracy function; and the duration corresponding to the minimum torque accuracy value is used as the time bias constant.

[0140] The angle data calculation module can also be used to determine the maximum angle data and the minimum angle data from the multiple historical angle data; determine the amplitude coefficient based on the maximum angle data and the minimum angle data; input the multiple historical angle data into a preset second loss function to calculate the phase coefficient; input the multiple historical angle data into a preset frequency function to calculate the frequency change coefficient; and establish the walking trajectory function based on the amplitude coefficient, the phase coefficient and the frequency change coefficient.

[0141] The angle data calculation module can also be used to obtain multiple historical walking angles on either side of the hip joint within a preset statistical time period before the corresponding moment of the current state; input the multiple historical walking angles into a preset normalization algorithm to generate normalized data corresponding to each of the historical walking angles; determine the historical angle data on either side of the hip joint in N gait cycles from the normalized data; N is a positive integer greater than or equal to 2.

[0142] The angle data calculation module can also be used to input multiple normalized data and the collection time corresponding to each normalized data into a preset cycle algorithm to determine the gait cycle duration; based on the gait cycle duration, determine the historical angle data of any side of the hip joint in the N gait cycles from the normalized data.

[0143] As for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method embodiment part.

[0144] Reference Figure 7 , shows a schematic diagram of a terminal device provided in an embodiment of the present application. Figure 7 As shown, the terminal device 700 in the embodiment of the present application includes: a processor 710, a memory 720, and a computer program 721 stored in the memory 720 and executable on the processor 710. When the processor 710 executes the computer program 721, the steps in each embodiment of the above-mentioned method for determining the hip joint assistance value are implemented, for example Figure 1 Alternatively, when the processor 710 executes the computer program 721, the functions of each module / unit in the above-mentioned device embodiments are realized, for example Figure 6 Functions of modules 601 to 602 are shown.

[0145] Exemplarily, the computer program 721 may be divided into one or more modules / units, which are stored in the memory 720 and executed by the processor 710 to complete the present application. The one or more modules / units may be a series of computer program instruction segments capable of completing specific functions, which may be used to describe the execution process of the computer program 721 in the terminal device 700. For example, the computer program 721 may be divided into an angle data calculation module and an assist value determination module, and the specific functions of each module are as follows:

[0146] An angle data calculation module, for calculating the first angle data of the user based on a walking trajectory function and a preset time offset constant in response to the power assistance calculation task on either side of the hip joint in the current state; the walking trajectory function is constructed by at least two collected historical angle data;

[0147] The assist value determination module is used to determine the assist value of either side of the hip joint in the current state based on the first angle data and the second angle data of either side of the hip joint in the current state.

[0148] The terminal device 700 may be a terminal device in each of the above embodiments. The terminal device 700 may include, but is not limited to, a processor 710 and a memory 720. Those skilled in the art will appreciate that Figure 7 It is only an example of the terminal device 700 and does not constitute a limitation of the terminal device 700. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the terminal device 700 may also include input and output devices, network access devices, buses, etc.

[0149] The processor 710 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0150] The memory 720 may be an internal storage unit of the terminal device 700, such as a hard disk or memory of the terminal device 700. The memory 720 may also be an external storage device of the terminal device 700, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the terminal device 700. Further, the memory 720 may also include both an internal storage unit of the terminal device 700 and an external storage device. The memory 720 is used to store the computer program 721 and other programs and data required by the terminal device 700. The memory 720 may also be used to temporarily store data that has been output or is to be output.

[0151] An embodiment of the present application also discloses a terminal device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the method for determining the hip joint assistance value as described in the aforementioned embodiments is implemented.

[0152] The embodiment of the present application further discloses a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for determining the hip joint assistance value as described in the above-mentioned embodiments is implemented.

[0153] The embodiment of the present application also discloses a computer program product. When the computer program product is run on a computer, the computer is enabled to execute the method for determining the hip joint assistance value described in the aforementioned embodiments.

[0154] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application is described in detail with reference to the above-mentioned embodiments, a person skilled in the art should understand that the technical solutions described in the above-mentioned embodiments can still be modified, or some of the technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A method for determining a hip joint assistance value, characterized in that: include: In response to the power calculation task on either side of the hip joint in the current state, first angle data of the user is calculated based on a walking trajectory function and a preset time offset constant; the walking trajectory function is constructed by at least two collected historical angle data; Determine the power assist value of either side of the hip joint in the current state based on the first angle data and the second angle data of either side of the hip joint in the current state; The determining the power assist value of either side of the hip joint in the current state based on the first angle data and the second angle data of either side of the hip joint in the current state comprises: determining an angle difference based on the first angle data and the second angle data; Inputting the angle difference into a torque estimation model to determine a torque estimation value; Determining the assist value based on a preset gain coefficient and the torque estimation value; Before inputting the angle difference into the torque estimation model to determine the torque estimation value, the method further includes: Obtaining a model to be trained and a training sample input by a user, and processing the training sample through the model to be trained to generate an initial torque value; the training sample includes a plurality of training angle values ​​and an expected torque value corresponding to each training angle value; Determine a loss value between the initial torque value and the expected torque value based on a preset first loss function; The model to be trained is updated based on the loss value until the loss value satisfies a preset training stop condition, and the model to be trained corresponding to when the loss value satisfies the training stop condition is used as the torque estimation model; After updating the model to be trained based on the loss value until the loss value satisfies a preset training stop condition, and using the model to be trained corresponding to when the loss value satisfies the training stop condition as the moment estimation model, the method further includes: Inputting the torque estimation model and the training sample into a preset evaluation function to generate a torque accuracy function; the torque accuracy function is used to determine the corresponding relationship between torque and duration; Solving the minimum torque for the torque accurate function to determine the minimum torque accurate value corresponding to the torque accurate function; The duration corresponding to the minimum torque accurate value is used as the time bias constant; In response to the power assistance calculation task on either side of the hip joint in the current state, before calculating the first angle data of the user based on the walking trajectory function and the preset time offset constant, the method includes: Determine maximum angle data and minimum angle data from the plurality of historical angle data; determining an amplitude coefficient based on the maximum angle data and the minimum angle data; Inputting a plurality of historical angle data into a preset second loss function to calculate a phase coefficient; Inputting the plurality of historical angle data into a preset frequency function to calculate a frequency variation coefficient; The walking trajectory function is established based on the amplitude coefficient, the phase coefficient and the frequency change coefficient.

2. The method according to claim 1, characterized in that Before determining the maximum angle data and the minimum angle data from the plurality of historical angle data, the method further comprises: Acquire multiple historical walking angles of any side of the hip joint within a preset statistical time before the corresponding moment of the current state; Inputting the multiple historical walking angles into a preset normalization algorithm to generate normalized data corresponding to each of the historical walking angles; The historical angle data of either side of the hip joint in N gait cycles is determined from the normalized data; wherein N is a positive integer greater than or equal to 2.

3. The method according to claim 2, characterized in that Determining a plurality of the historical angle data of any side of the hip joint in N gait cycles from the normalized data comprises: Inputting the plurality of normalized data and the collection time corresponding to each of the normalized data into a preset cycle algorithm to determine the duration of the gait cycle; The historical angle data of either side of the hip joint in the N gait cycles are determined from the normalized data based on the gait cycle duration.

4. A device for determining a hip joint assistance value, characterized in that: include: An angle data calculation module, for calculating the first angle data of the user based on a walking trajectory function and a preset time offset constant in response to the power assistance calculation task on either side of the hip joint in the current state; the walking trajectory function is constructed by at least two collected historical angle data; an assist value determination module, configured to determine an assist value of either side of the hip joint in a current state based on the first angle data and the second angle data of either side of the hip joint in a current state; The determining the power assist value of either side of the hip joint in the current state based on the first angle data and the second angle data of either side of the hip joint in the current state comprises: determining an angle difference based on the first angle data and the second angle data; Inputting the angle difference into a torque estimation model to determine a torque estimation value; Determining the assist value based on a preset gain coefficient and the torque estimation value; Before inputting the angle difference into the torque estimation model to determine the torque estimation value, the method further includes: Obtaining a model to be trained and a training sample input by a user, and processing the training sample through the model to be trained to generate an initial torque value; the training sample includes a plurality of training angle values ​​and an expected torque value corresponding to each training angle value; Determine a loss value between the initial torque value and the expected torque value based on a preset first loss function; The model to be trained is updated based on the loss value until the loss value satisfies a preset training stop condition, and the model to be trained corresponding to when the loss value satisfies the training stop condition is used as the torque estimation model; After updating the model to be trained based on the loss value until the loss value satisfies a preset training stop condition, and using the model to be trained corresponding to when the loss value satisfies the training stop condition as the moment estimation model, the method further includes: Inputting the torque estimation model and the training sample into a preset evaluation function to generate a torque accuracy function; the torque accuracy function is used to determine the corresponding relationship between torque and duration; Solving the minimum torque for the torque accurate function to determine the minimum torque accurate value corresponding to the torque accurate function; The duration corresponding to the minimum torque accurate value is used as the time bias constant; In response to the power assistance calculation task on either side of the hip joint in the current state, before calculating the first angle data of the user based on the walking trajectory function and the preset time offset constant, the method includes: Determine maximum angle data and minimum angle data from the plurality of historical angle data; determining an amplitude coefficient based on the maximum angle data and the minimum angle data; Inputting a plurality of historical angle data into a preset second loss function to calculate a phase coefficient; Inputting the plurality of historical angle data into a preset frequency function to calculate a frequency variation coefficient; The walking trajectory function is established based on the amplitude coefficient, the phase coefficient and the frequency change coefficient.

5. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method for determining the hip joint assistance value as described in any one of claims 1-3 is implemented.

6. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for determining the hip joint assistance value as described in any one of claims 1-3 is implemented.

Images