A real-time recognition method for human motion patterns in a flexible exoskeleton system

Abstract

The present invention discloses a method for real-time recognition of human motion patterns of a flexible exoskeleton system, comprising the following steps: the flexible exoskeleton system acquires human posture information and human-ground interaction force information in real time; sets the initial state of the human motion pattern of the flexible exoskeleton system at the initial moment; the flexible exoskeleton system determines the human motion pattern that may be switched at the current moment based on the human motion pattern at the previous moment, and then determines the recognition conditions for switching the human motion pattern based on the human posture signal and human-ground interaction force signal acquired at the current moment, and finally obtains the recognition result of the human motion pattern at the current moment. The present invention involves using the state switching model and the real-time measurement data of the exoskeleton system sensor to determine whether the state recognition conditions of the human motion pattern are established, and obtains the human motion pattern recognition result in real time. The method is simple, easy to apply, and there is no time delay in the recognition result, thus realizing the engineering application of human motion pattern recognition.

Description

Technical Field

[0001] The present invention relates to the technical field of exoskeleton systems, and in particular to a real-time recognition method for human motion patterns of a flexible exoskeleton system. Background Art

[0002] The exoskeleton system is a typical human-machine intelligent system that combines the machine's ability to bear loads with the advantages of the human body's free guidance. Therefore, it requires good coordination between humans and machines to achieve efficient assistance while ensuring the comfort of human movement.

[0003] Human motion pattern recognition technology is one of the key technologies of exoskeleton technology. It plays a key role in the coordinated control and flexible assistance of the exoskeleton system. Its characteristics are to calculate and estimate the human motion pattern based on human posture information, human-machine interaction force information and human-ground interaction force information. The main motion patterns of people in daily life include standing, squatting, walking, running and jumping. For different motion patterns, the human body's muscle work mode and bone force transmission mode are different. Therefore, real-time recognition of human motion patterns is of great significance to the wearing comfort and walking flexibility of the exoskeleton system.

[0004] Traditional exoskeleton systems are highly rigid. They obtain models by machine learning training of human-computer interaction force information and human posture information, and then use online data to match the models to obtain pattern classification. However, for flexible exoskeleton systems, in order to improve wearing comfort and walking flexibility, the joints are designed with flexible connection technology, and human-computer interaction force information cannot be obtained. In addition, machine learning methods require learning and training of sample data. For different people, the movement postures are different, and the required sample volume is huge, which is difficult to collect. In addition, when performing pattern matching online, it is necessary to judge the data of a time period, the amount of data calculation is large, and the pattern matching results have a time delay, which cannot meet the real-time requirements of the exoskeleton system. In addition, the complexity of machine learning methods is large, and a high-demand hardware platform is required. The application difficulty and cost are very high.

[0005] In general, the real-time recognition method of human motion patterns in traditional exoskeleton systems is difficult to obtain data. Different people may have different motion postures, exercise habits, physical conditions, etc. The workload of collecting sample data is huge, and it is difficult to be complete for different people; the real-time nature of the recognition results is difficult to guarantee, and the recognition results will always be delayed for a period of time, making it difficult to guarantee their real-time nature; the algorithms are complex and difficult to apply. Machine learning algorithms are usually difficult, with huge amounts of calculations and algorithm complexity. The requirements for the hardware platform during the application process are very high, which brings inconvenience to engineering applications. Summary of the invention

[0006] Based on the above problems, the purpose of the present invention is to provide a real-time recognition method for human motion patterns of a flexible exoskeleton system, which mainly solves the technical problem that the human motion pattern recognition method of the traditional exoskeleton system is not suitable for human motion pattern recognition of flexible exoskeletons.

[0007] In order to solve the above problems, the present invention provides the following technical solutions:

[0008] The present invention provides a method for real-time recognition of human motion patterns in a flexible exoskeleton system, characterized in that it comprises the following steps:

[0009] S1, the flexible exoskeleton system obtains human posture information and human-ground interaction force information in real time;

[0010] S2, setting the starting state of the human body motion mode of the flexible exoskeleton system at the initial moment;

[0011] S3, the flexible exoskeleton system determines the human motion mode that can be switched at the current moment according to the human motion mode at the previous moment, then determines the recognition condition of the human motion mode switching according to the human posture signal and human-ground interaction force signal obtained at the current moment, and finally obtains the result of the human motion mode at the current moment;

[0012] S4, repeat the previous step S3 to obtain the recognition result of the human body movement pattern at each moment until the system operation ends.

[0013] Furthermore, an IMU inertial measurement device for acquiring the human body posture signal is installed on the back, thigh, calf and foot of the flexible exoskeleton system, and a pressure sensor for acquiring the human-ground interaction force signal is installed on the foot of the flexible exoskeleton system.

[0014] Furthermore, the human body posture signal includes the joint angles of the human body's hip, knee, and ankle joints and the posture angle of the foot.

[0015] Furthermore, the human body movement modes include standing mode, walking assistance mode, jumping mode, squatting mode, running mode and transition mode.

[0016] Furthermore, the switching of the human body motion mode includes six situations:

[0017] Q1, the previous moment is the standing mode, and the flexible exoskeleton system determines the recognition conditions of the squatting mode, the jumping mode and the standing mode respectively according to the human posture signal and the human-ground interaction force signal acquired at the current moment;

[0018] Q2, the previous moment is the walking assistance mode, and the flexible exoskeleton system randomly determines the recognition conditions of the standing mode, the jumping mode and the running mode according to the human posture signal and the human-ground interaction force signal obtained at the current moment;

[0019] Q3, the previous moment is the jumping mode, and the flexible exoskeleton system determines the recognition condition of the standing mode according to the human posture signal and human-ground interaction force signal acquired at the current moment;

[0020] Q4, the previous moment is the squatting mode, and the flexible exoskeleton system randomly determines the recognition conditions of the standing mode and the jumping mode according to the human posture signal and the human-ground interaction force signal obtained at the current moment;

[0021] Q5, the previous moment is the running mode, and the flexible exoskeleton system determines the recognition conditions of the standing mode, the jumping mode and the running mode in sequence according to the human posture signal and the human-ground interaction force signal acquired at the current moment;

[0022] Q6, the previous moment is the transition mode, and the flexible exoskeleton system determines the recognition conditions of the squatting mode, the jumping mode, the standing mode and the running mode according to the human posture signal and the human-ground interaction force signal obtained at the current moment.

[0023] Furthermore, the recognition condition of the standing mode is:

[0024] a. One of the two feet of the flexible exoskeleton system is determined to be supported on the ground, that is, the human-ground interaction force signal of the left or right foot is required to be greater than the set foot pressure threshold;

[0025] b. The change in the plantar IMU measurement value of the flexible exoskeleton system over a period of time is less than the set foot posture angle threshold;

[0026] c. The knee joint angle of the flexible exoskeleton system is less than the set knee joint angle threshold;

[0027] If the three conditions are met simultaneously, the recognition condition of the standing mode is met.

[0028] Furthermore, the recognition condition of the squat mode is:

[0029] a. The knee joint angles of both legs of the flexible exoskeleton system are simultaneously greater than the set knee joint angle threshold, and the duration is greater than the set time threshold;

[0030] b. At least one of the two feet of the flexible exoskeleton system is determined to be supported on the ground, that is, the human-ground interaction force signal of the left foot or the right foot is required to be greater than the set foot pressure threshold and the duration is greater than the set time threshold;

[0031] c. the previous moment was the standing mode or transition mode;

[0032] If the three conditions are met simultaneously, the recognition condition of the standing mode is met.

[0033] Furthermore, the recognition condition of the skip mode is:

[0034] Both feet of the flexible exoskeleton system leave the ground at the same time, that is, the human-machine interaction force signals of the left foot and the right foot are simultaneously less than the set foot pressure threshold and the duration is greater than the set time threshold;

[0035] If the condition is satisfied, the recognition condition of the skip mode is reached.

[0036] Furthermore, the recognition condition of the running mode is:

[0037] a. The last moment is the standing mode, transition mode or walking assistance mode;

[0038] b. Detecting that the duration of the support state within a gait cycle is less than a set time threshold;

[0039] If both conditions are met, the recognition condition of the running mode is met.

[0040] Furthermore, the identification condition of the transition mode is:

[0041] a. The last moment is the standing mode;

[0042] b. At the current moment, one of the two feet of the flexible exoskeleton system leaves the ground, that is, the human-ground interaction force signal of the left foot or the right foot is required to be less than the set foot pressure threshold.

[0043] If both conditions are satisfied at the same time, the recognition condition of the transition mode is met.

[0044] Furthermore, when the recognition conditions of the walking assistance mode: the standing mode, transition mode, squat mode, jump mode, and running mode are all not met, it is determined to be the walking assistance mode.

[0045] Furthermore, the processor of the flexible exoskeleton system performs smoothing filtering on the human posture signal and the human-ground interaction force signal to eliminate the influence of noise and jitter.

[0046] Compared with the prior art, the present invention provides the following beneficial effects:

[0047] The present invention relates to a real-time recognition method for human motion patterns of a flexible exoskeleton system. The method utilizes a state switching algorithm and real-time measurement data of an exoskeleton system sensor to determine whether a state recognition condition for a human motion pattern is established, and obtains a human motion pattern recognition result in real time. The method is simple, easy to apply, and has no time delay in the recognition result, thereby realizing the engineering application of human motion pattern recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0049] Figure 1 It is the overall structure diagram of the flexible exoskeleton system of the present invention;

[0050] Figure 2 This is a flow chart of human motion pattern recognition of the flexible exoskeleton system of the present invention;

[0051] Figure 3 This is a flow chart of switching the human body motion mode when the previous moment is the standing mode of the present invention;

[0052] Figure 4 This is a flow chart of switching the human body motion mode when the present invention is in the walking assistance mode at the last moment;

[0053] Figure 5 The present invention is a flow chart of switching the human body motion mode when the previous moment is the jumping mode;

[0054] Figure 6 The present invention is a flow chart of switching the human body motion mode when the previous moment is in the squatting mode;

[0055] Figure 7 The present invention is a flow chart of switching the human body motion mode when the previous moment is the running mode;

[0056] Figure 8 The present invention is a flow chart of switching the human body motion mode when the previous moment is a transition mode;

[0057] Reference numerals:

[0058] 1. Back; 2. Thigh; 3. Calf; 4. Foot; 5. Bowden cable; 6. Back frame; 7. Strap; 8. Drive system. DETAILED DESCRIPTION

[0059] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0060] like Figures 1 to 2 As shown, this embodiment discloses a method for real-time recognition of human motion patterns of a flexible exoskeleton system, the method comprising the following steps:

[0061] S1, the flexible exoskeleton system obtains human posture information and human-ground interaction force information in real time;

[0062] S2, setting the starting state of the human body motion mode of the flexible exoskeleton system at the initial moment;

[0063] S3, the flexible exoskeleton system determines the human motion mode that can be switched at the current moment according to the human motion mode at the previous moment, then determines the recognition condition of the human motion mode switching according to the human posture signal and human-ground interaction force signal obtained at the current moment, and finally obtains the result of the human motion mode at the current moment;

[0064] S4, repeat the previous step S3 to obtain the recognition result of the human body movement pattern at each moment until the system operation ends.

[0065] like Figure 1 As shown, the driving system 8 of the flexible exoskeleton system is installed on the back frame 6, and is connected to the calf 3 of the flexible exoskeleton system through the Bowden cable 5. A part of the Bowden cable 5 is passed through the Bowden cable sleeve, one end of the Bowden cable sleeve is fixed on the driving system 8, and the other end is fixed on the thigh 2. The driving system 8 achieves assistance and unloading of the knee joint by retracting and releasing the Bowden cable 5, thereby achieving soft and coordinated assistance control of the human body.

[0066] The tension sensor is installed on the calf 3 of the flexible exoskeleton system to measure the torque output feedback of the drive system 8. The angular displacement sensor is installed inside the drive system 8 to measure the position output feedback of the flexible exoskeleton system; the IMU inertial measurement device is installed on the back 1, thigh 2, calf 3 and foot 4 of the flexible exoskeleton system to obtain human posture information, and the pressure sensor is installed on the foot 4 of the flexible exoskeleton system to obtain human-ground interaction force information.

[0067] Regarding the installation method of the above-mentioned IMU inertial measurement device, specifically, the direction of a certain axis of the IMU inertial measurement device is the same as the sagittal axis of the human body, and in the embodiment, it is the X-axis direction. The angle measurement range of the IMU inertial measurement device is 0 to 360°. The human body posture signal includes the angle values ​​of the human body's hip, knee, and ankle joints and the posture angle of the foot. Specifically, the above-mentioned joint angles are defined as 90°, the knee joint angle is defined as 0°, and the ankle joint angle is defined as 90° in the standing state. The hip joint angle increases when the thigh swings forward, and the knee joint angle increases when the knee is bent, the toes are tilted up, the ankle joint angle is closed, and the foot posture angle decreases. The above-mentioned foot posture angle is specifically 180° in the standing state, the toes are lifted up, the angle increases, and the toes are pressed down, and the angle decreases.

[0068] IMU inertial measurement devices and pressure sensors are installed on the exoskeleton system and tied to the human body through flexible weaving technology. They follow the human body's movements and measure the human body's posture information and human-ground interaction force information in real time. The detected signals are sent to the processor of the flexible exoskeleton system through bus transmission. The processor performs smooth filtering and angle calculation on these signals. The processor of the flexible exoskeleton system performs smooth filtering on the human body posture signal and human-ground interaction force signal to eliminate the influence of noise.

[0069] Under different motion mode conditions, the human body has different posture angle information and human-ground interaction force. The processor of the flexible exoskeleton system extracts motion mode features and switches modes based on the detected human posture signals and human-ground interaction force signals.

[0070] The human motion mode includes standing mode, transition mode, walking assistance mode, squatting mode, jumping mode, and running mode. The walking assistance mode includes walking on flat ground, going up and down stairs, and going up and down slopes. In this embodiment, when the system first determines the state switching of the human motion mode, the initial state of the human motion mode at the initial moment is set to standing mode.

[0071] like Figures 3 to 8 As shown in the figure, the switching of human motion mode includes six situations:

[0072] Q1, such as Figure 3 As shown, the previous moment was the standing mode, and the flexible exoskeleton system determines the recognition conditions of the squatting mode, the jumping mode, and the standing mode in turn according to the human posture signal and the human-ground interaction force signal obtained at the current moment.

[0073] The determination order is: first determine the recognition conditions of the squat mode, if they are met, switch to the squat mode; if not, then determine the recognition conditions of the jump mode, if they are met, switch to the jump mode; if not, finally determine the recognition conditions of the standing mode, if they are met, continue to maintain the standing mode, if not, switch to the transition mode.

[0074] Q2, such as Figure 4 As shown, the previous moment was the walking assistance mode, and the flexible exoskeleton system randomly determines the recognition conditions of the standing mode, the jumping mode and the running mode according to the human posture signal and the human-ground interaction force signal obtained at the current moment.

[0075] There is no clear requirement on the determination order. In the present embodiment, the determination order is as follows: first determine the recognition conditions of the standing mode. If they are met, switch to the standing mode. If they are not met, then determine the recognition conditions of the jumping mode. If they are met, switch to the jumping mode. If they are not met, finally determine the recognition conditions of the running mode. If they are met, switch to the running mode. If they are not met, maintain the walking assistance mode.

[0076] Q3, such as Figure 5 As shown, the previous moment was in jumping mode. The flexible exoskeleton system determines the standing mode recognition conditions based on the human posture signal and human-ground interaction force signal obtained at the current moment. If the conditions are met, it switches to standing mode. If not, it maintains jumping mode.

[0077] Q4, such as Figure 6 As shown, the previous moment was the squatting mode, and the flexible exoskeleton system randomly determines the standing mode and jumping mode recognition conditions according to the human posture signal and human-ground interaction force signal obtained at the current moment.

[0078] There is no clear requirement for the order of recognition. In this embodiment, the order is: first determine the recognition conditions of the standing mode. If they are met, switch to the standing mode. If they are not met, then determine the recognition conditions of the jumping mode. If they are met, switch to the jumping mode. If they are not met, continue to maintain the squatting mode.

[0079] Q5, such as Figure 7 As shown, the previous moment was the running mode. The exoskeleton system determines the recognition conditions of the standing mode, the jumping mode and the running mode respectively according to the human body posture signal and the human-ground interaction force signal obtained at the current moment.

[0080] The order of determination is: first determine the recognition conditions of the standing mode. If they are met, switch to the standing mode. If not, then determine the recognition conditions of the jumping mode. If they are met, switch to the jumping mode. If not, finally determine the recognition conditions of the running mode. If they are not met, switch to the walking assist mode. If they are met, remain in the running mode.

[0081] Q6, such as Figure 8 As shown, the previous moment is the transition mode. The flexible exoskeleton system randomly determines the recognition conditions of the squat mode, the jump mode, the standing mode, and the running mode according to the human posture signal and the human-ground interaction force signal obtained at the current moment.

[0082] There is no clear requirement for the recognition order. The determination order in the embodiment is: first determine the recognition conditions of the squatting mode. If they are met, switch to the squatting mode; if they are not met, determine the recognition conditions of the jumping mode. If they are met, switch to the jumping mode; if they are not met, determine the recognition conditions of the standing mode. If they are met, switch to the standing mode; if they are not met, determine the recognition conditions of the running mode. If they are met, switch to the running mode; if they are not met, switch to the walking assist mode.

[0083] The recognition conditions for standing mode are:

[0084] a. One of the two feet of the flexible exoskeleton system is determined to be supported on the ground, that is, the human-ground interaction force signal of the left foot or the right foot is required to be greater than the set foot pressure threshold. In this embodiment, the foot pressure threshold is 3 kg.

[0085] b. The change of the IMU measurement value of the flexible exoskeleton system in a period of time is less than the set foot posture angle threshold, that is, the foot posture is required to have no obvious change in a period of time. In this embodiment, the change of the IMU measurement value of the left foot and the right foot is less than 10 degrees.

[0086] c. The knee joint angle of the flexible exoskeleton system is less than the set knee joint angle threshold. In this embodiment, the left and right knee joint angles are both less than 15 degrees.

[0087] The above three conditions need to be met simultaneously to meet the recognition conditions of the standing mode.

[0088] The recognition conditions for squat mode are:

[0089] a. The knee joint angles of both legs of the flexible exoskeleton system are greater than the set knee joint angle threshold at the same time, and the duration is greater than the set time threshold, that is, both legs are required to maintain a flexed knee state at the same time. In this embodiment, the left and right knee joint angles are both greater than 45 degrees, and the duration is greater than 50 milliseconds.

[0090] b. At least one of the two feet of the flexible exoskeleton system is determined to be supported on the ground, and the duration is greater than the set time threshold. In this embodiment, the measurement value of the plantar pressure sensor of the left or right foot is greater than 3 kilograms, and the duration is greater than 50 milliseconds.

[0091] c. The last moment was standing mode or transition mode, and other modes cannot be switched.

[0092] If the three conditions are met simultaneously, the recognition condition of the standing mode is met.

[0093] The recognition conditions of the jumping mode are as follows: both feet of the flexible exoskeleton system leave the ground at the same time, that is, the human-computer interaction force signals of the left foot and the right foot are simultaneously less than the set foot pressure threshold, and the duration is greater than the set time threshold. In this embodiment, the plantar pressure test value of the left foot or the right foot is less than 3 kg.

[0094] The recognition conditions of running mode are:

[0095] a. The last moment is the standing mode, transition mode or walking assistance mode, and other modes are not switched.

[0096] b. Detect that the duration of the support state in one gait cycle is less than a set time threshold, that is, the duration of the support state in one gait cycle is less than 0.35 seconds.

[0097] The identification conditions of the transition mode are:

[0098] a. The last moment was standing mode.

[0099] b. At the current moment, one of the two feet of the flexible exoskeleton system leaves the ground, that is, the human-ground interaction force signal of the left foot or the right foot is required to be less than the set foot pressure threshold. In this embodiment, the plantar pressure of the left foot or the right foot is less than 3 kilograms.

[0100] The recognition conditions of the walking assistance mode are: when the recognition conditions of the standing mode, transition mode, squatting mode, jumping mode and running mode are all not met, it is determined to be the walking assistance mode.

[0101] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for real-time recognition of human motion patterns in a flexible exoskeleton system, characterized in that: The following steps are involved: S1, the flexible exoskeleton system obtains human posture information and human-ground interaction force information in real time; S2, setting the starting state of the human body motion mode of the flexible exoskeleton system at the initial moment; S3, the flexible exoskeleton system determines the human motion mode that can be switched at the current moment according to the human motion mode at the previous moment, then determines the recognition condition of the human motion mode switching according to the acquired human posture information and human-ground interaction force information at the current moment, and finally obtains the result of the human motion mode at the current moment; S4, repeating the previous step S3 to obtain the recognition result of the human body movement pattern at each moment until the system operation ends; The human body posture information includes the joint angles of the human body's hip, knee, ankle and foot posture angles; The human body motion modes include standing mode, transition mode, walking assistance mode, squatting mode, jumping mode and running mode; The switching of the human body motion mode includes: Q1, the previous moment is the standing mode, and the flexible exoskeleton system determines the recognition conditions of the squatting mode, the jumping mode and the standing mode in sequence according to the human posture information and the human-ground interaction force information obtained at the current moment; Q2, the previous moment is the walking assistance mode, and the flexible exoskeleton system randomly determines the recognition conditions of the standing mode, the jumping mode and the running mode according to the human posture information and the human-ground interaction force information obtained at the current moment; Q3, the previous moment is the jumping mode, and the flexible exoskeleton system determines the recognition condition of the standing mode according to the human posture information and human-ground interaction force information obtained at the current moment; Q4, the previous moment is the squatting mode, and the flexible exoskeleton system randomly determines the recognition conditions of the standing mode and the jumping mode according to the human posture information and human-ground interaction force information obtained at the current moment; Q5, the previous moment is the running mode, and the flexible exoskeleton system determines the recognition conditions of the standing mode, the jumping mode and the running mode in sequence according to the human posture information and the human-ground interaction force information obtained at the current moment; Q6, the previous moment is the transition mode, and the flexible exoskeleton system randomly determines the recognition conditions of the squat mode, the recognition conditions of the jump mode, the recognition conditions of the stand mode and the recognition conditions of the run mode according to the human posture information and the human-ground interaction force information obtained at the current moment; The drive system of the flexible exoskeleton system is installed on the back frame and connected to the calf of the flexible exoskeleton system through a Bowden cable. A part of the Bowden cable is inserted into a Bowden cable sleeve. One end of the Bowden cable sleeve is fixed to the drive system, and the other end is fixed to the thigh. The drive system realizes the assistance and unloading of the knee joint by retracting and releasing the Bowden cable, thereby realizing the soft and coordinated assistance control of the human body. The tension sensor is installed on the calf of the flexible exoskeleton system to measure the torque output feedback of the drive system; the angular displacement sensor is installed inside the drive system to measure the position output feedback of the flexible exoskeleton system; the IMU inertial measurement device is installed on the back, thigh, calf and foot of the flexible exoskeleton system to obtain human posture information, and the pressure sensor is installed on the foot of the flexible exoskeleton system to obtain human-ground interaction force information; Among them, the direction of one axis of the IMU inertial measurement device is the same as the sagittal axis of the human body, which is the X-axis direction, and the angle measurement range of the IMU inertial measurement device is 0 to 360°; the human posture information includes the angle values ​​of the human hip, knee, and ankle joints and the posture angle of the foot; the joint angle, in the standing state, the hip joint angle is defined as 90°, the knee joint angle is defined as 0°, and the ankle joint angle is defined as 90°. The hip joint angle increases when the thigh swings forward, and the knee joint angle increases when the knee is bent, the toes are raised, the ankle joint angle is closed, and the foot posture angle decreases; the foot posture angle is 180° in the standing state, the toes are raised, the angle increases, and the toes are pressed down, the angle decreases.

2. The method for real-time recognition of human motion patterns of a flexible exoskeleton system according to claim 1, characterized in that: An IMU inertial measurement device for acquiring the human body posture information is installed on the back, thigh, calf and foot of the flexible exoskeleton system, and a pressure sensor for acquiring the human-ground interaction force information is installed on the foot of the flexible exoskeleton system.

3. The method for real-time recognition of human motion patterns of a flexible exoskeleton system according to claim 1, characterized in that: The recognition conditions of the standing mode are: a. One of the two feet of the flexible exoskeleton system is determined to be supported on the ground, that is, the human-ground interaction force information of the left or right foot is required to be greater than the set foot pressure threshold; b. The change in the plantar IMU measurement value of the flexible exoskeleton system over a period of time is less than the set foot posture angle threshold; c. The knee joint angle of the flexible exoskeleton system is less than the set knee joint angle threshold; If the three conditions are met simultaneously, the recognition condition of the standing mode is met.

4. The method for real-time recognition of human motion patterns of a flexible exoskeleton system according to claim 1, characterized in that: The recognition conditions of the squat mode are: a. The knee joint angles of both legs of the flexible exoskeleton system are simultaneously greater than the set knee joint angle threshold, and the duration is greater than the set time threshold; b. At least one of the two feet of the flexible exoskeleton system is determined to be supported on the ground, that is, the human-ground interaction force information of the left foot or the right foot is required to be greater than the set foot pressure threshold and the duration is greater than the set time threshold; c. the previous moment was the standing mode or transition mode; If the three conditions are met simultaneously, the recognition condition of the standing mode is met.

5. The method for real-time recognition of human motion patterns of a flexible exoskeleton system according to claim 1, characterized in that: The recognition condition of the skip mode is: Both feet of the flexible exoskeleton system leave the ground at the same time, that is, the human-machine interaction force information of the left foot and the right foot is simultaneously less than the set foot pressure threshold, and the duration is greater than the set time threshold; If the condition is satisfied, the recognition condition of the skip mode is reached.

6. The method for real-time recognition of human motion patterns in a flexible exoskeleton system according to claim 1, characterized in that: The recognition conditions of the running mode are: a. The last moment is the standing mode, transition mode or walking assistance mode; b. Detecting that the duration of the support state within a gait cycle is less than a set time threshold; If both conditions are met, the recognition condition of the running mode is met.

7. The method for real-time recognition of human motion patterns in a flexible exoskeleton system according to claim 1, characterized in that: The identification conditions of the transition mode are: a. The last moment is the standing mode; b. One of the two feet of the flexible exoskeleton system at the current moment leaves the ground, that is, the human-ground interaction force information of the left foot or the right foot is required to be less than the set foot pressure threshold; If both conditions are satisfied at the same time, the recognition condition of the transition mode is met.

8. The method for real-time recognition of human motion patterns in a flexible exoskeleton system according to claim 1, characterized in that: The recognition conditions of the walking assistance mode: when the recognition conditions of the standing mode, transition mode, squatting mode, jumping mode and running mode are all not met, it is determined to be the walking assistance mode.

9. The method for real-time recognition of human motion patterns in a flexible exoskeleton system according to any one of claims 1 to 8, characterized in that: The processor of the flexible exoskeleton system performs smoothing filtering on the human posture information and human-ground interaction force information to eliminate the influence of noise and jitter.

Images