A method and system for gait parameter adaptive optimization based on LSTM

Abstract

The application provides a gait parameter adaptive optimization method and system based on LSTM, which comprises the following steps: constructing a training set, wherein the training set comprises gait analysis data and joint angles of rescue personnel in various environments and gait modes on the well, and the gait analysis data comprises inertial measurement signals, electromyographic signals and plantar pressure signals; constructing a prediction model based on LSTM, training the prediction model by using the training set to obtain a trained prediction model; obtaining real-time gait analysis data of the current time step of the rescue personnel in the environment underground, inputting the real-time gait analysis data into the trained prediction model to output an initial value of the joint angle prediction of the next time step, combining the expected joint angle to enhance the electromyographic signals in the real-time gait analysis data, and then reusing the trained prediction model to output a target value of the joint angle prediction of the next time step, so as to adaptively adjust the gait parameters of the equipment worn by the rescue personnel based on the target value of the joint angle prediction in the next time step.

Description

Technical Field

[0001] The present invention relates to the field of gait analysis and optimization, and in particular to a gait parameter adaptive optimization method and system based on LSTM. Background Art

[0002] After an accident in a mine, emergency rescue is required. In the complex and ever-changing emergency rescue environment, the equipment worn by rescuers needs to adjust the gait parameters in real time based on the collected electromyographic signals according to the continuous action mode to improve the efficiency and safety of the action. Traditional gait adjustment and optimization methods rely on preset algorithms or manual adjustments, which cannot meet the needs of fast and complex rescue tasks. In addition, underground air, dust pollution, human body temperature, magnetic field and other environments have a great impact on the intensity of the collected electromyographic signals, resulting in inaccurate gait adjustment, which in turn affects the rescue efficiency. Therefore, it is necessary to develop a gait parameter optimization system that can automatically learn and adapt to different rescue scenarios based on the differences in individual characteristics and environmental characteristics, and realize continuous prediction of human movement within a stable gait cycle. Summary of the Invention

[0003] The present invention aims to solve one of the technical problems in the related art at least to a certain extent.

[0004] To this end, the first object of the present invention is to propose a gait parameter adaptive optimization method based on LSTM to perform gait adjustment in a timely and accurate manner.

[0005] The second object of the present invention is to propose a gait parameter adaptive optimization system based on LSTM.

[0006] A third object of the present invention is to provide an electronic device.

[0007] A fourth object of the present invention is to provide a computer-readable storage medium.

[0008] To achieve the above objectives, the first aspect of the present invention proposes a gait parameter adaptive optimization method based on LSTM, comprising:

[0009] Constructing a training set, wherein the training set includes gait analysis data and joint angles of multiple historical time steps of rescuers in various environments and gait modes on the well, wherein the gait analysis data includes inertial measurement signals, electromyographic signals, and plantar pressure signals;

[0010] Constructing a prediction model based on LSTM, wherein the input of the prediction model is gait analysis data of the time step before the target time step, and the output is the joint angle prediction value of the target time step, and the prediction model is trained using the training set to obtain a trained prediction model;

[0011] Underground, real-time gait analysis data of the current time step of the rescuer's environment is obtained, and the real-time gait analysis data is input into a trained prediction model to output an initial value of the joint angle prediction for the next time step after the current time step;

[0012] enhancing the electromyographic signal in the real-time gait analysis data based on the joint angle prediction initial value and the corresponding expected joint angle to obtain optimized real-time gait analysis data for the current time step;

[0013] The optimized real-time gait analysis data is input into a trained prediction model to output a joint angle prediction target value for the next time step of the current time step, so as to adaptively adjust the gait parameters of the equipment worn by the rescuer based on the joint angle prediction target value at the next time step.

[0014] In the method of the first aspect of the present invention, the electromyographic signals include electromyographic signals of multiple muscle positions of the rescuer, and the multiple muscle positions include the rectus femoris, vastus medialis, and vastus lateralis muscles related to knee extension movement, the semitendinosus and biceps femoris muscles related to knee flexion movement, the tibialis anterior muscle related to ankle dorsiflexion movement, and the medial gastrocnemius, lateral gastrocnemius, and soleus muscles related to ankle plantar flexion movement.

[0015] In the method of the first aspect of the present invention, the output of the prediction model also includes a gait pattern recognition result. When underground, the optimized real-time gait analysis data is input into the trained prediction model, which also outputs the gait pattern recognition result of the next time step of the current time step. At the next time step, the gait parameters of the equipment worn by the rescue personnel are adaptively adjusted based on the joint angle prediction target value and the gait pattern recognition result.

[0016] In the method of the first aspect of the present invention, the process of obtaining the expected joint angle includes: after obtaining the trained prediction model, collecting gait analysis data of the rescue personnel in different environments and gait patterns on the well; for the collected gait analysis data of the rescue personnel in different environments and gait patterns on the well, respectively use the trained prediction model to obtain the corresponding joint angle prediction value, and the joint angle prediction value obtained for the different environments and gait patterns on the well is the expected joint angle in the corresponding environment, wherein each expected joint angle and the corresponding environment, gait pattern and data features of the gait analysis data constitute a mapping group, and the data features of the gait analysis data are obtained by the feature extractor in the trained prediction model.

[0017] In the method of the first aspect of the present invention, the process of determining the expected joint angle corresponding to the initial value of the joint angle prediction includes: using the feature extractor in the trained prediction model to extract features from the real-time gait analysis data to obtain data features of the real-time gait analysis data; determining the corresponding mapping group based on the data features of the real-time gait analysis data and the environment in which the rescue personnel are located, and the expected joint angle in the mapping group is the expected joint angle corresponding to the initial value of the joint angle prediction.

[0018] In the method of the first aspect of the present invention, the enhancing of the electromyographic signal in the real-time gait analysis data based on the joint angle prediction initial value and the corresponding expected joint angle includes: calculating the error between the joint angle prediction initial value and the corresponding expected joint angle; determining a compensation matrix based on the error and the environment in which the rescuer is located; and multiplying the electromyographic signal in the real-time gait analysis data by the compensation matrix to achieve the enhancement of the electromyographic signal in the real-time gait analysis data.

[0019] To achieve the above objectives, the second aspect of the present invention proposes a gait parameter adaptive optimization system based on LSTM, comprising:

[0020] A training set construction module is used to construct a training set, wherein the training set includes gait analysis data and joint angles of multiple historical time steps of rescuers in various environments and gait modes on the well, wherein the gait analysis data includes inertial measurement signals, electromyographic signals, and plantar pressure signals;

[0021] A modeling module is used to build a prediction model based on LSTM, wherein the input of the prediction model is the gait analysis data of the time step before the target time step, and the output is the joint angle prediction value of the target time step, and the prediction model is trained using the training set to obtain a trained prediction model;

[0022] An acquisition module is used to obtain real-time gait analysis data of the current time step of the rescuer's environment underground, input the real-time gait analysis data into a trained prediction model, and output an initial value of the joint angle prediction for the next time step after the current time step;

[0023] an enhancement module, configured to enhance the electromyographic signal in the real-time gait analysis data based on the joint angle prediction initial value and the corresponding expected joint angle, so as to obtain optimized real-time gait analysis data for a current time step;

[0024] The adaptive adjustment module is used to input the optimized real-time gait analysis data into a trained prediction model to output the joint angle prediction target value of the next time step of the current time step, so as to adaptively adjust the gait parameters of the equipment worn by the rescuer based on the joint angle prediction target value at the next time step.

[0025] In the system of the second aspect of the present invention, the output of the prediction model in the modeling module also includes a gait pattern recognition result. In the adaptive adjustment module, the optimized real-time gait analysis data is input into the trained prediction model and the gait pattern recognition result of the next time step of the current time step is also output. At the next time step, the gait parameters of the equipment worn by the rescuer are adaptively adjusted based on the joint angle prediction target value and the gait pattern recognition result.

[0026] To achieve the above-mentioned purpose, the third aspect of the present invention proposes an electronic device, comprising: a processor, and a memory communicatively connected to the processor; the memory stores computer-executable instructions; the processor executes the computer-executable instructions stored in the memory to implement the method proposed in the first aspect of the present invention.

[0027] To achieve the above-mentioned purpose, the fourth aspect of the present invention proposes a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the method proposed in the first aspect of the present invention.

[0028] The LSTM-based gait parameter adaptive optimization method, system, electronic device and storage medium provided by the present invention construct a training set, which includes gait analysis data and joint angles of multiple historical time steps of rescue personnel in various environments and gait modes on the well, and the gait analysis data includes inertial measurement signals, electromyographic signals and plantar pressure signals; a prediction model is constructed based on LSTM, the input of the prediction model is the gait analysis data of the time step before the target time step, and the output is the joint angle prediction value of the target time step, and the prediction model is trained using the training set to obtain a trained prediction model; underground, the rescue personnel's environment is obtained, and the training set is used to train the prediction model. The real-time gait analysis data of the current time step in the environment is input into the trained prediction model to output the initial value of the joint angle prediction for the next time step of the current time step; the electromyographic signal in the real-time gait analysis data is enhanced based on the initial value of the joint angle prediction and the corresponding expected joint angle to obtain the optimized real-time gait analysis data of the current time step; the optimized real-time gait analysis data is input into the trained prediction model to output the target value of the joint angle prediction for the next time step of the current time step, so as to adaptively adjust the gait parameters of the equipment worn by the rescuer based on the target value of the joint angle prediction at the next time step. In this case, a prediction model is constructed based on LSTM, and a training set is constructed using gait analysis data and joint angles of multiple historical time steps in various environments and gait modes on the well to train the prediction model. When underground, the influence of the underground environment on the electromyographic signal is taken into consideration. After obtaining the initial value of the joint angle prediction, the electromyographic signal in the real-time gait analysis data is enhanced by combining the initial value of the joint angle prediction and the corresponding expected joint angle to reduce the influence of the underground environment on the electromyographic signal. The trained prediction model is then used to obtain the target value of the joint angle prediction. The target value of the joint angle prediction is more accurate than the initial value of the joint angle prediction, thereby achieving timely and accurate gait adjustment when the next time step arrives.

[0029] Additional aspects and advantages of the present invention will be set forth in part in the description which follows and, in part, will be obvious from the description which follows, or may be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0031] Figure 1 A schematic diagram of a flow chart of a gait parameter adaptive optimization method based on LSTM provided in an embodiment of the present invention;

[0032] Figure 2 A block diagram of an LSTM-based gait parameter adaptive optimization system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0033] The following describes embodiments of the present invention in detail, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and are not to be construed as limiting the present invention.

[0034] The following describes a method and system for adaptive optimization of gait parameters based on LSTM (Long Short-Term Memory) according to an embodiment of the present invention with reference to the accompanying drawings.

[0035] The embodiment of the present invention provides a gait parameter adaptive optimization method based on LSTM to perform gait adjustment in a timely and accurate manner.

[0036] Figure 1 A flowchart of an LSTM-based gait parameter adaptive optimization method provided in an embodiment of the present invention.

[0037] like Figure 1 As shown in FIG, the LSTM-based gait parameter adaptive optimization method includes the following steps:

[0038] Step S101: construct a training set, which includes gait analysis data and joint angles of multiple historical time steps of rescuers in various environments and gait modes on the well. The gait analysis data includes inertial measurement signals, electromyographic signals, and plantar pressure signals.

[0039] In step S101 , the various environments include, for example, different air environments, dust environments, magnetic field environments, and the like.

[0040] In step S101 , the gait pattern includes but is not limited to walking, jogging, going up and down stairs, going up and down slopes, carrying, dragging, and other motion states.

[0041] In step S101 , each historical time step is at least greater than or equal to 500 ms.

[0042] In step S101, gait analysis data includes inertial measurement unit (IMU) signals, electromyographic (EMG) signals, and plantar pressure signals. The EMG signals include EMG signals from multiple muscle locations of the rescuer, including but not limited to the rectus femoris, vastus medialis, and vastus lateralis muscles associated with knee extension, the semitendinosus and biceps femoris muscles associated with knee flexion, the tibialis anterior muscles associated with ankle dorsiflexion, and the medial, lateral, and soleus muscles associated with ankle plantar flexion.

[0043] In step S101, inertial measurement signals (IMU), electromyographic signals (EMG) and plantar pressure signals can be collected by sensors at corresponding positions in the equipment worn by the rescuer.

[0044] In step S101, when constructing the training set, the acquired gait analysis data may also be pre-processed (such as noise reduction) and invalid value elimination, etc., to improve the accuracy of the model during subsequent training.

[0045] Step S102: construct a prediction model based on LSTM. The input of the prediction model is the gait analysis data of the time step before the target time step, and the output is the joint angle prediction value of the target time step. The prediction model is trained using the training set to obtain a trained prediction model.

[0046] In step S102, it is easy to understand that the Long Short-Term Memory (LSTM) network is a time-recurrent neural network that is widely used in time series prediction problems.

[0047] In step S102, a prediction model is constructed based on the LSTM. The prediction model is used to sequentially extract features and predict angles from the input data. Specifically, the prediction model includes a feature extractor and an angle predictor, thereby performing feature extraction and angle prediction on the input data.

[0048] In step S102, the prediction model is fed with gait analysis data from the time step preceding the target time step. Specifically, the prediction model includes the inertial measurement unit (IMU), electromyography (EMG), and plantar pressure signals from the time step preceding the target time step. The target time step can be any time step, such as any historical time step, the current time step, or the time step immediately following the current time step.

[0049] In step S102 , the output of the prediction model is the predicted value of the joint angle at the target time step.

[0050] In step S102, the output of the prediction model also includes the gait pattern recognition result at the target time step. The gait pattern recognition result is the predicted movement state of the rescuer, including walking, trotting, climbing stairs, climbing up and down slopes, carrying, dragging, and other movement states.

[0051] In step S102, the prediction model is trained using the training set to obtain a trained prediction model. In embodiments of the present invention, the trained prediction model is capable of identifying motion states in various environments, such as walking, jogging, climbing stairs, ascending and descending slopes, carrying, and dragging. In embodiments of the present invention, the trained prediction model has a motion state recognition accuracy rate of ≥94%; an average error in continuous joint angle prediction of ≤5°; and a motion state lead prediction time of ≥500ms.

[0052] In step S102, after obtaining the trained prediction model on the well, it is also necessary to obtain the expected joint angle so that the expected joint angle can be directly retrieved when needed underground in subsequent steps.

[0053] In step S102, the process of obtaining the expected joint angle includes: after obtaining the trained prediction model, collecting gait analysis data of the rescue personnel in different environments and gait patterns on the well; for the collected gait analysis data of the rescue personnel in different environments and gait patterns on the well, respectively use the trained prediction model to obtain corresponding joint angle prediction values, and the joint angle prediction values ​​obtained for different environments and gait patterns on the well are the expected joint angles in the corresponding environments, wherein each expected joint angle and the corresponding environment, gait pattern and data features of the gait analysis data constitute a mapping group, and the data features of the gait analysis data are obtained by the feature extractor in the trained prediction model.

[0054] Step S103: Underground, obtain real-time gait analysis data of the current time step in the environment where the rescuer is located, input the real-time gait analysis data into the trained prediction model to output the initial value of the joint angle prediction of the next time step of the current time step.

[0055] In step S103, the real-time gait analysis data includes real-time inertial measurement signals, electromyographic signals, and plantar pressure signals. The real-time inertial measurement signals, electromyographic signals, and plantar pressure signals of the current time step are input into a trained prediction model, which outputs the joint angle prediction value for the next time step after the current time step. Considering the impact of the underground environment on the real-time electromyographic signals, which reduces the accuracy of the joint angle prediction value to a certain extent, the joint angle prediction value in this step is the initial joint angle prediction value, and subsequent enhancement processing is required to improve the prediction accuracy.

[0056] In step S103, the real-time gait analysis data is input into the trained prediction model, and the gait pattern recognition result of the next time step of the current time step is output.

[0057] Step S104 : enhancing the electromyographic signal in the real-time gait analysis data based on the predicted initial value of the joint angle and the corresponding expected joint angle, so as to obtain optimized real-time gait analysis data of the current time step.

[0058] In step S104, the process of determining the expected joint angle corresponding to the initial value of the joint angle prediction includes: using the feature extractor in the trained prediction model to extract features from the real-time gait analysis data to obtain data features of the real-time gait analysis data; determining the corresponding mapping group based on the data features of the real-time gait analysis data and the environment in which the rescue personnel are located, and the expected joint angle in the mapping group is the expected joint angle corresponding to the initial value of the joint angle prediction.

[0059] In step S104, the myoelectric signals in the real-time gait analysis data are enhanced based on the initial joint angle prediction values ​​and the corresponding desired joint angles. This includes calculating the error between the initial joint angle prediction values ​​and the corresponding desired joint angles; determining a compensation matrix based on the error and the rescuer's environment; and multiplying the myoelectric signals in the real-time gait analysis data by the compensation matrix to enhance the myoelectric signals in the real-time gait analysis data. Thus, error backpropagation optimizes the model input data and enhances the accuracy of the model prediction.

[0060] In step S104 , the enhanced electromyographic signal of the current time step, the inertial measurement signal and the plantar pressure signal in real time with the current time step constitute the optimized real-time gait analysis data of the current time step.

[0061] In step S105, the optimized real-time gait analysis data is input into the trained prediction model to output the joint angle prediction target value for the next time step of the current time step, so as to adaptively adjust the gait parameters of the equipment worn by the rescuer based on the joint angle prediction target value at the next time step.

[0062] In step S105, the optimized real-time gait analysis data is input into the trained prediction model to output the predicted target value of the joint angle for the next time step of the current time step.

[0063] In step S105, while underground, the optimized real-time gait analysis data is input into the trained prediction model, which also outputs the gait pattern recognition results for the time step following the current time step. The enhancement of the electromyographic signals does not affect the predicted gait pattern recognition results, so the gait pattern recognition results in this step are consistent with those output in step S103.

[0064] In step S105, when the next time step arrives, the gait parameters (i.e., joint angles and gait patterns) of the equipment worn by the rescuer are adaptively adjusted based on the optimal gait parameters (i.e., joint angle prediction target values ​​and gait pattern recognition results).

[0065] In some embodiments, the method of the present invention can achieve adaptive learning optimization of wearable equipment for rescuers with a height of, for example, 1.6 to 1.9 meters and a weight of, for example, 50 to 90 kilograms. The method of the present invention further includes: determining a mapping relationship between height, weight, and expected joint angles; after determining the expected joint angle corresponding to the initial value of the joint angle prediction, correcting the obtained expected joint angle based on the mapping relationship and the height and weight of the current rescuer to obtain an expected joint angle correction value; enhancing the electromyographic signal in the real-time gait analysis data based on the initial value of the joint angle prediction and the corresponding expected joint angle correction value, thereby obtaining a joint angle prediction target value and a gait pattern recognition result.

[0066] In order to implement the above embodiment, the present invention also proposes a gait parameter adaptive optimization system based on LSTM.

[0067] Figure 2 A block diagram of an LSTM-based gait parameter adaptive optimization system provided in an embodiment of the present invention.

[0068] like Figure 2 As shown, the LSTM-based gait parameter adaptive optimization system includes a training set construction module 11, a modeling module 12, an acquisition module 13, an enhancement module 14 and an adaptive adjustment module 15, wherein:

[0069] A training set construction module 11 is used to construct a training set, the training set including gait analysis data and joint angles of multiple historical time steps of rescuers in various environments and gait modes on the well, the gait analysis data including inertial measurement signals, electromyographic signals and plantar pressure signals;

[0070] A modeling module 12 is used to build a prediction model based on LSTM. The input of the prediction model is the gait analysis data of the time step before the target time step, and the output is the joint angle prediction value of the target time step. The prediction model is trained using the training set to obtain a trained prediction model.

[0071] The acquisition module 13 is used to obtain real-time gait analysis data of the current time step of the rescuer's environment underground, input the real-time gait analysis data into the trained prediction model, and output the initial value of the joint angle prediction for the next time step after the current time step;

[0072] an enhancement module 14 for enhancing the electromyographic signal in the real-time gait analysis data based on the predicted initial value of the joint angle and the corresponding expected joint angle to obtain optimized real-time gait analysis data for the current time step;

[0073] The adaptive adjustment module 15 is used to input the optimized real-time gait analysis data into the trained prediction model to output the joint angle prediction target value of the next time step of the current time step, so as to adaptively adjust the gait parameters of the equipment worn by the rescuer based on the joint angle prediction target value at the next time step.

[0074] Furthermore, in a possible implementation of an embodiment of the present invention, in the training set construction module 11, the electromyographic signals include electromyographic signals of multiple muscle positions of the rescuer, and the multiple muscle positions include the rectus femoris, vastus medialis, and vastus lateralis muscles related to knee extension movement, the semitendinosus and biceps femoris muscles related to knee flexion movement, the tibialis anterior muscle related to ankle dorsiflexion movement, and the medial gastrocnemius, lateral gastrocnemius, and soleus muscles related to ankle plantar flexion movement.

[0075] Furthermore, in a possible implementation of an embodiment of the present invention, the output of the prediction model in the modeling module 12 also includes a gait pattern recognition result. In the adaptive adjustment module 15, the optimized real-time gait analysis data is input into the trained prediction model and the gait pattern recognition result of the next time step of the current time step is also output. At the next time step, the gait parameters of the equipment worn by the rescuer are adaptively adjusted based on the joint angle prediction target value and the gait pattern recognition result.

[0076] Furthermore, in a possible implementation of an embodiment of the present invention, the LSTM-based gait parameter adaptive optimization system also includes an expected joint angle acquisition module, which is connected to the modeling module 12 and the enhancement module 14 respectively, and the expected joint angle acquisition module is used to collect gait analysis data of rescue personnel in different environments and gait patterns on the well; for the collected gait analysis data of rescue personnel in different environments and gait patterns on the well, the trained prediction model is used to obtain corresponding joint angle prediction values, and the joint angle prediction values ​​obtained for different environments and gait patterns on the well are the expected joint angles in the corresponding environments, wherein each expected joint angle constitutes a mapping group with the corresponding environment, gait pattern and data features of the gait analysis data, and the data features of the gait analysis data are obtained by the feature extractor in the trained prediction model.

[0077] Furthermore, in a possible implementation of an embodiment of the present invention, in the enhancement module 14, the process of determining the expected joint angle corresponding to the initial value of the joint angle prediction includes: using the feature extractor in the trained prediction model to extract features from the real-time gait analysis data to obtain data features of the real-time gait analysis data; determining a corresponding mapping group based on the data features of the real-time gait analysis data and the environment in which the rescue personnel are located, and the expected joint angle in the mapping group is the expected joint angle corresponding to the initial value of the joint angle prediction.

[0078] Furthermore, in a possible implementation of an embodiment of the present invention, in the enhancement module 14, the electromyographic signal in the real-time gait analysis data is enhanced based on the joint angle prediction initial value and the corresponding expected joint angle, including: calculating the error between the joint angle prediction initial value and the corresponding expected joint angle; determining a compensation matrix based on the error and the environment in which the rescuer is located; and multiplying the electromyographic signal in the real-time gait analysis data by the compensation matrix to achieve enhancement of the electromyographic signal in the real-time gait analysis data.

[0079] It should be noted that the above explanation of the embodiment of the LSTM-based gait parameter adaptive optimization method is also applicable to the LSTM-based gait parameter adaptive optimization system of this embodiment, and will not be repeated here.

[0080] In an embodiment of the present invention, a training set is constructed, which includes gait analysis data and joint angles of rescue personnel in multiple historical time steps in various environments and gait modes on the well, and the gait analysis data includes inertial measurement signals, electromyographic signals, and plantar pressure signals; a prediction model is constructed based on LSTM, the input of the prediction model is the gait analysis data of the time step before the target time step, and the output is the joint angle prediction value of the target time step, and the prediction model is trained using the training set to obtain a trained prediction model; underground, real-time gait analysis data of the current time step in the environment where the rescue personnel are located is obtained, the real-time gait analysis data is input into the trained prediction model to output the initial value of the joint angle prediction for the next time step of the current time step; the electromyographic signal in the real-time gait analysis data is enhanced based on the initial value of the joint angle prediction and the corresponding expected joint angle to obtain optimized real-time gait analysis data for the current time step; the optimized real-time gait analysis data is input into the trained prediction model to output the target value of the joint angle prediction for the next time step of the current time step, so that the gait parameters of the equipment worn by the rescue personnel can be adaptively adjusted based on the target value of the joint angle prediction at the next time step. In this case, a prediction model is constructed based on LSTM, and a training set is constructed using gait analysis data and joint angles of multiple historical time steps in various environments and gait modes on the well to train the prediction model. When underground, the influence of the underground environment on the electromyographic signal is taken into consideration. After obtaining the initial value of the joint angle prediction, the electromyographic signal in the real-time gait analysis data is enhanced by combining the initial value of the joint angle prediction and the corresponding expected joint angle to reduce the influence of the underground environment on the electromyographic signal. The trained prediction model is then used to obtain the target value of the joint angle prediction. The target value of the joint angle prediction is more accurate than the initial value of the joint angle prediction, thereby achieving timely and accurate gait adjustment when the next time step arrives.

[0081] The method and system of the present invention are a LSTM-based adaptive optimization method and system for gait parameters of emergency rescue wearable equipment. The method and system are applied to adaptively adjust the gait parameters of wearable equipment in complex rescue environments. By collecting gait analysis data of rescuers while performing rescue missions, the method uses a prediction model to analyze and predict the optimal gait parameters under different environments, and automatically adjusts the wearable equipment to achieve optimal performance. Compared with the existing technology, the present invention has the following advantages: (1) It can timely and accurately realize dynamic adaptive optimization of gait parameters, improving the mobility, efficiency, and safety of rescuers; and (2) It improves the overall rescue efficiency and the use effect of wearable equipment.

[0082] In order to implement the above embodiments, the present invention also proposes an electronic device, comprising: a processor, and a memory communicatively connected to the processor; the memory stores computer-executable instructions; the processor executes the computer-executable instructions stored in the memory to implement the method provided by the above embodiments.

[0083] In order to implement the above embodiments, the present invention further provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the methods provided in the above embodiments.

[0084] In order to implement the above embodiments, the present invention further provides a computer program product, including a computer program, which implements the methods provided in the above embodiments when executed by a processor.

[0085] The collection, storage, use, processing, transmission, provision and disclosure of user personal information involved in this invention are in compliance with the relevant laws and regulations and do not violate public order and good morals.

[0086] It is important to note that personal information collected from users should be used for legitimate and reasonable purposes and should not be shared or sold beyond these legitimate uses. Furthermore, such collection / sharing should be conducted only after receiving the user's informed consent, including but not limited to notifying the user to read the user agreement / user notice and sign an agreement / authorization that includes the relevant user information before using the feature. Furthermore, any necessary steps must be taken to safeguard and secure access to such personal information and ensure that others with access to personal information comply with its privacy policy and procedures.

[0087] The present invention contemplates providing implementations that allow users to selectively block the use or access of personal information data. Specifically, the present disclosure contemplates providing hardware and / or software to prevent or block access to such personal information data. Risks can be minimized by limiting data collection and deleting data once it is no longer needed. Furthermore, where applicable, such personal information can be de-identified to protect user privacy.

[0088] In the descriptions of the foregoing embodiments, the reference terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" mean that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification and features of different embodiments or examples, unless they are mutually inconsistent.

[0089] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present invention, "plurality" means at least two, such as two, three, etc., unless otherwise specifically defined.

[0090] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may be performed out of the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present invention pertain.

[0091] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic devices, and a portable compact disc read-only memory (CDROM). Furthermore, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing it in another suitable manner if necessary, and then storing it in a computer memory.

[0092] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0093] Those skilled in the art will understand that all or part of the steps in the method of the above embodiment can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0094] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing module, or each unit may exist physically separately, or two or more units may be integrated into a single module. The aforementioned integrated modules may be implemented in the form of hardware or in the form of software functional modules. If the integrated modules are implemented in the form of software functional modules and sold or used as independent products, they may also be stored in a computer-readable storage medium.

[0095] The storage medium mentioned above may be a read-only memory, a magnetic disk, or an optical disk, etc. Although the embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and are not to be construed as limiting the present invention. Persons skilled in the art may make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

Claims

1. A gait parameter adaptive optimization method based on LSTM, characterized in that: include: Constructing a training set, wherein the training set includes gait analysis data and joint angles of multiple historical time steps of rescuers in various environments and gait modes on the well, wherein the gait analysis data includes inertial measurement signals, electromyographic signals, and plantar pressure signals; Constructing a prediction model based on LSTM, wherein the input of the prediction model is gait analysis data of the time step before the target time step, and the output is the joint angle prediction value of the target time step, and the prediction model is trained using the training set to obtain a trained prediction model; Underground, real-time gait analysis data of the current time step of the rescuer's environment is obtained, and the real-time gait analysis data is input into a trained prediction model to output an initial value of the joint angle prediction for the next time step after the current time step; The electromyographic signal in the real-time gait analysis data is enhanced based on the joint angle prediction initial value and the corresponding expected joint angle to obtain the optimized real-time gait analysis data of the current time step, wherein the process of determining the expected joint angle corresponding to the joint angle prediction initial value includes: using the feature extractor in the trained prediction model to extract features from the real-time gait analysis data to obtain data features of the real-time gait analysis data, determining a corresponding mapping group based on the data features of the real-time gait analysis data and the environment in which the rescuer is located, the expected joint angle in the mapping group being the expected joint angle corresponding to the joint angle prediction initial value; enhancing the electromyographic signal in the real-time gait analysis data based on the joint angle prediction initial value and the corresponding expected joint angle, including: calculating the error between the joint angle prediction initial value and the corresponding expected joint angle; determining a compensation matrix based on the error and the environment in which the rescuer is located, and multiplying the electromyographic signal in the real-time gait analysis data by the compensation matrix to achieve enhancement of the electromyographic signal in the real-time gait analysis data; The optimized real-time gait analysis data is input into a trained prediction model to output a joint angle prediction target value for the next time step of the current time step, so as to adaptively adjust the gait parameters of the equipment worn by the rescuer based on the joint angle prediction target value at the next time step.

2. The LSTM-based gait parameter adaptive optimization method according to claim 1, characterized in that: The electromyographic signals include electromyographic signals of multiple muscle positions of the rescuer, including the rectus femoris, vastus medialis, and vastus lateralis muscles related to knee extension movement, the semitendinosus and biceps femoris muscles related to knee flexion movement, the tibialis anterior muscle related to ankle dorsiflexion movement, and the medial gastrocnemius, lateral gastrocnemius, and soleus muscles related to ankle plantar flexion movement.

3. The LSTM-based gait parameter adaptive optimization method according to claim 1, characterized in that: The output of the prediction model also includes gait pattern recognition results. When underground, the optimized real-time gait analysis data is input into the trained prediction model, which also outputs the gait pattern recognition results of the next time step of the current time step. At the next time step, the gait parameters of the equipment worn by the rescue personnel are adaptively adjusted based on the joint angle prediction target value and the gait pattern recognition results.

4. The LSTM-based gait parameter adaptive optimization method according to claim 1, characterized in that: The process of obtaining the desired joint angle includes: After obtaining the trained prediction model, gait analysis data of rescuers in different environments and gait patterns on the well were collected; For the gait analysis data collected from rescue personnel in different environments and gait patterns on the well, the trained prediction model is used to obtain the corresponding joint angle prediction values. The joint angle prediction values ​​obtained for different environments and gait patterns on the well are the expected joint angles in the corresponding environments, where each expected joint angle constitutes a mapping group with the corresponding environment, gait pattern and data features of the gait analysis data. The data features of the gait analysis data are obtained by the feature extractor in the trained prediction model.

5. The LSTM-based gait parameter adaptive optimization method according to claim 1, characterized in that: The process of determining the expected joint angle corresponding to the initial value of the joint angle prediction includes: Using a feature extractor in a trained prediction model to perform feature extraction on the real-time gait analysis data to obtain data features of the real-time gait analysis data; A corresponding mapping group is determined based on the data features of the real-time gait analysis data and the environment in which the rescuer is located. The expected joint angle in the mapping group is the expected joint angle corresponding to the initial value of the joint angle prediction.

6. The LSTM-based gait parameter adaptive optimization method according to claim 1, characterized in that: The step of enhancing the electromyographic signal in the real-time gait analysis data based on the predicted initial value of the joint angle and the corresponding expected joint angle comprises: Calculate the error between the initial value of the predicted joint angle and the corresponding expected joint angle; Determining a compensation matrix based on the error and the rescuer's environment; The myoelectric signal in the real-time gait analysis data is multiplied by the compensation matrix to enhance the myoelectric signal in the real-time gait analysis data.

7. A gait parameter adaptive optimization system based on LSTM, characterized in that: include: A training set construction module is used to construct a training set, wherein the training set includes gait analysis data and joint angles of multiple historical time steps of rescuers in various environments and gait modes on the well, wherein the gait analysis data includes inertial measurement signals, electromyographic signals, and plantar pressure signals; A modeling module is used to build a prediction model based on LSTM, wherein the input of the prediction model is the gait analysis data of the time step before the target time step, and the output is the joint angle prediction value of the target time step, and the prediction model is trained using the training set to obtain a trained prediction model; An acquisition module is used to obtain real-time gait analysis data of the current time step of the rescuer's environment underground, input the real-time gait analysis data into a trained prediction model, and output an initial value of the joint angle prediction for the next time step after the current time step; An enhancement module is used to enhance the electromyographic signal in the real-time gait analysis data based on the joint angle prediction initial value and the corresponding expected joint angle to obtain optimized real-time gait analysis data of the current time step, wherein the process of determining the expected joint angle corresponding to the joint angle prediction initial value includes: using a feature extractor in a trained prediction model to extract features from the real-time gait analysis data to obtain data features of the real-time gait analysis data, determining a corresponding mapping group based on the data features of the real-time gait analysis data and the environment in which the rescuer is located, the expected joint angle in the mapping group being the expected joint angle corresponding to the joint angle prediction initial value; enhancing the electromyographic signal in the real-time gait analysis data based on the joint angle prediction initial value and the corresponding expected joint angle includes: calculating an error between the joint angle prediction initial value and the corresponding expected joint angle; determining a compensation matrix based on the error and the environment in which the rescuer is located, and multiplying the electromyographic signal in the real-time gait analysis data by the compensation matrix to achieve enhancement of the electromyographic signal in the real-time gait analysis data; The adaptive adjustment module is used to input the optimized real-time gait analysis data into a trained prediction model to output the joint angle prediction target value of the next time step of the current time step, so as to adaptively adjust the gait parameters of the equipment worn by the rescuer based on the joint angle prediction target value at the next time step.

8. The LSTM-based gait parameter adaptive optimization system according to claim 7, characterized in that: The output of the prediction model in the modeling module also includes a gait pattern recognition result. In the adaptive adjustment module, the optimized real-time gait analysis data is input into the trained prediction model and the gait pattern recognition result of the next time step of the current time step is also output. At the next time step, the gait parameters of the equipment worn by the rescuer are adaptively adjusted based on the joint angle prediction target value and the gait pattern recognition result.

9. An electronic device, characterized in that: include: a processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the method according to any one of claims 1 to 6.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 6 when executed by a processor.

Images