Customized multi-objective optimization design method suitable for wearable energy harvesting assisting equipment

Abstract

The invention discloses a customized multi-objective optimization design method suitable for wearable energy harvesting power assisting equipment, and relates to the field of wearable power assisting equipment. The problem that in the prior art, the daily assistance or training effect cannot be achieved on the premise that the energy recovery efficiency is maximized is solved. Recording weight of the target wearer, lower limb kinematics in a single gait cycle and plantar pressure data; calculating joint power data; determining the power and assistance output model of the equipment according to the selected configuration of the energy harvesting assistance equipment; setting an electric energy generation power optimization target and an equipment assistance effect optimization target; setting equipment power constraint, spring moment constraint and stay wire tension constraint; establishing a multi-objective optimization design problem, performing optimization calculation, and obtaining an optimization design Pareto frontier; according to user requirements, customized optimal design parameters are selected on the Pareto frontier; and adjusting the energy harvesting equipment to obtain a customized optimal auxiliary effect. The method is used for optimizing the wearable power assisting equipment.

Description

Technical Field

[0001] The present invention relates to the field of wearable assistive devices, and particularly to a customized multi-objective optimization design method applicable to wearable energy harvesting assistive devices, for achieving an optimized balance between the assistive effect and the power generation efficiency. Background Art

[0002] With the intensification of the aging population situation in China, the demand for wearable daily physiological monitoring devices is increasing day by day. However, commonly used daily physiological monitoring devices often consume a large amount of energy and have a high demand for power supply batteries, greatly increasing the weight burden on the wearer. In recent years, the emergence of wearable energy harvesting devices has effectively alleviated this problem. It recovers the energy of the wearer's joints and converts it into electrical energy to provide power for wearable daily physiological monitoring devices. However, existing wearable energy harvesting devices often have relatively single functions, are difficult to adapt to the physiological characteristics and gait habits of different wearers, and it is difficult to achieve the optimal state of energy recovery efficiency.

[0003] For example: the invention patent with the publication number CN 107126348 B and the patent name of a quasi-passive knee-ankle coupled lower limb exoskeleton and its control method. The control method of this exoskeleton is as follows: measure the sole pressure through the sole pressure sensor in the control mechanism and output a resistance signal regarding the sole pressure, collect the resistance signal through a sampling circuit, the sampling circuit converts the resistance signal into a voltage signal, and outputs the voltage signal to a single-chip microcomputer; the single-chip microcomputer receives the voltage signal, determines the gait cycle of human walking according to the voltage signal; the single-chip microcomputer outputs PWM signals to two servo motors according to the gait cycle to respectively control the rotation angles of the power output shafts of the two servo motors; according to the rotation of the power output shafts of the two servo motors, respectively control the opening or locking of the knee joint clutch and the ankle joint clutch.

[0004] It controls the actions in different stages by dividing the gait cycle. However, this division is mainly for different stages of the total swing during the human walking process, such as: storing energy, assisting, etc. However, it cannot provide the optimal design parameters of a customized wearable energy harvesting assistive device according to the joint power characteristics of different wearers, achieve an optimized balance between the assistive effect and the power generation efficiency, improve the overall operation effect of such devices, and optimize the daily experience of users. Summary of the Invention

[0005] The object of the present invention is to solve the problem that existing wearable energy-harvesting assistive devices cannot achieve daily assistance or training effects on the premise of maximizing the energy recovery efficiency (maximizing the energy recovery efficiency means that when a human walks, the lower limb joints have two states of acceleration and deceleration. When accelerating, positive work can be applied to the joints to assist their movement, and when decelerating, negative work can be applied to the joints to help them decelerate. This device uses the damping force of the motor to apply decelerating negative work to the joints to help them decelerate, and at the same time converts the absorbed negative work into electrical energy. The ratio of the recovered electric power to the negative power of the human joint can be defined as the energy recovery efficiency. The negative power of the human joint cannot be directly measured. The present invention believes that the change in the human body's metabolic level can characterize the effect of applying negative work assistance to the joints. Therefore, the energy recovery efficiency here can be defined as the ratio of the change in the metabolic level to the generated electric power.), and then provide a customized multi-objective optimization design method applicable to wearable energy-harvesting assistive devices.

[0006] The technical solution of the present invention is as follows:

[0007] A customized multi-objective optimization design method applicable to wearable energy-harvesting assistive devices, comprising the following steps:

[0008] Step 1: Record the weight of the target wearer, the lower limb kinematics and plantar pressure data within a single gait cycle;

[0009] Among them, a single gait cycle of the wearer can be divided into n sampling points; the kinematic data within a single gait cycle are three one-dimensional arrays of length n, and each one-dimensional array includes the angle θ(n), angular velocity ω(n) and angular acceleration α(n) of each joint of the wearer's lower limb, as well as the linear acceleration Acc(n) of each limb segment of the lower limb, with the units being rad, rad / s, rad / s 2 and m / s 2 body;

[0010] Step 2: Calculate the joint power data of the target wearer based on the weight, lower limb kinematic data, and plantar pressure data;

[0011] Among them, calculating the joint power data of the target wearer includes the following steps:

[0012] Step 2-1: Consider the single leg of the wearer participating in data collection as a planar serial three-link model composed of three joints of the hip, knee, and ankle and three links of the thigh, calf, and foot;

[0013] Step 2-2: Use the dynamic calculation method, combined with the angle θ(n), angular velocity ω(n), angular acceleration α(n) of each joint, the linear acceleration Acc(n) of each limb segment, the plantar pressure data GRF(n), and the weight m of the wearer, to solve the joint torque data T(n) of each joint of the lower limb, with the unit being Nm;

[0014] Step 2-3: Based on the torque data T(n) of each lower limb joint and the angular velocity ω(n) of each lower limb joint, calculate the power P(n) of each lower limb joint, with the unit of W;

[0015] Step 3: According to the configuration of the selected energy-harvesting and assisting device, determine the power and assisting output model of the device;

[0016] Step 4: Set the optimization objectives for the power generated by electricity and the assisting effect of the device;

[0017] Step 5: Set the constraints on the device power, spring torque, and cable tension;

[0018] Step 6: Establish a multi-objective optimization design problem, conduct optimization calculations, and obtain the Pareto front of the optimization design. There are two opposite optimization design directions for the optimal assisting configuration and the optimal training configuration on this Pareto front; among them, the optimal assisting configuration aims to optimize the power generation efficiency while reducing the metabolic level of the wearer, and is suitable for users who need assistance in daily activities; the optimal training configuration aims to maximize the power generation efficiency while increasing the metabolic level of the wearer, and is suitable for users with daily physical training needs;

[0019] As described above, the multi-objective optimization design problem is specifically expressed as:

[0020] min f(x) = [f1(x), f2(x)] T

[0021] s.t. g j (x) ≤ 0, j = 1, 2, 3.

[0022] Among them, x is the decision vector for multi-objective optimization.

[0023] Step 7: According to the actual power generation demand or assisting demand of the user, select customized optimal design parameters on the Pareto front;

[0024] Step 8: Adjust the energy-harvesting device according to the selected parameters to obtain the customized optimal assisting effect.

[0025] Furthermore, the configuration of the selected energy-harvesting device in Step 3 is a device that is strapped to the outside of the lower limb of the wearer and uses a cable - pulley structure to collect the joint energy of the wearer and drive a generator to generate electricity. The radius of the pulley is r, with the unit of m; the energy-harvesting device has a spring energy storage mechanism for pre-tightening the cable and providing assisting force for movement, and the elastic coefficient of the spring energy storage mechanism is K, with the unit of Nm / rad.

[0026] Furthermore, the power generation power optimization target f1(x) in step four is quantified by the average electric power of the generator within a single gait cycle and is expressed as:

[0027]

[0028] In the formula, x is the decision vector of multi-objective optimization, E is the voltage generated by the device, and I is the current generated by the device.

[0029] Furthermore, the device assistance effect optimization target f2(x) in step four is measured by the change in the metabolic level of the wearer after being affected by the energy harvesting assistance device and is specifically expressed as:

[0030]

[0031] Among them, P + is the power of the positive work done by the energy harvesting device on the human body, and P - is the power of the negative work done by the energy harvesting device on the human body.

[0032] Furthermore, the device power constraint g1(x) in step five is specifically expressed as:

[0033]

[0034] Among them, P k is the biological power of the human knee joint, and P m is the power exerted by the energy harvesting device on the human body.

[0035] Furthermore, the spring moment constraint g2(x) and the cable tension constraint g3(x) in step five are to ensure that the cable is always in a pre-tensioned state, enabling the energy harvesting assistance device to always follow the movement of the wearer and maximizing the assistance efficiency, and are specifically expressed as:

[0036]

[0037] Among them, Ts is the torque of the scroll spring, and F is the cable tension.

[0038] Furthermore, the multi-objective optimization problem in step six is solved iteratively using a multi-objective optimization intelligent algorithm to obtain multiple sets of optimized design parameters that meet the requirements of each optimization target.

[0039] Furthermore, in step six, in the optimized design of the wearable knee joint energy harvesting assistance device, the decision vector x mentioned in the multi-objective optimization problem includes: the distance h1 from the fixed position of the energy harvesting assistance device on the wearer's thigh to the center of the wearer's knee joint, the distance h2 from the fixed position of the cable anchor point on the wearer's calf to the center of the wearer's knee joint, the radius r of the cable pulley, and the load resistance R of the energy harvesting circuit l, the elastic coefficient K of the elastic energy storage mechanism, and the spring pre-tightening angle θ0 of the elastic energy storage mechanism;

[0040] Specifically, it is expressed as:

[0041] x = [h1, h2, r, R l , K, θ0] T .

[0042] Furthermore, in step seven, according to the actual power generation demand or assistance demand of the user, customized optimal adjustment parameters are selected on the Pareto front, that is, according to the needs of the target user, the optimized design parameters of the wearable energy harvesting assistance device are selected from the optimal assistance configuration and the optimal training configuration. The optimized design parameters include: the distance h1 from the fixed position of the energy harvesting assistance device on the wearer's thigh to the center of the wearer's knee joint, the distance h2 from the fixed position of the cable anchor point on the wearer's calf to the center of the wearer's knee joint, the radius r of the cable pulley, and the load resistance R of the energy harvesting circuit l , the elastic coefficient K of the elastic energy storage mechanism, and the spring pre-tightening angle θ0 of the elastic energy storage mechanism.

[0043] Furthermore, in step eight, the energy harvesting device is adjusted according to the selected parameters to obtain a customized optimal assistance effect, that is, according to the optimized design parameters selected in step seven, the design and optimization adjustment of the energy harvesting assistance device are carried out for the target customer to meet the individual differences among different wearers.

[0044] The present invention has the following effects compared with the prior art:

[0045] Aiming at the joint power differences caused by different body types, walking habits, and joint lesion degrees of different target users, the present invention uses multi-objective optimization methods such as NSGAIII to calculate the Pareto front suitable for the user. Each point on the Pareto front represents an optimized design parameter. The user can, according to their specific needs, such as for daily training, the design parameters under the optimal training configuration will make the positive power cycle of the energy-enabling device show greater resistance during walking, while generating more electric energy, achieving the effect of daily rehabilitation training; if daily assistance is needed, the optimal assistance configuration is selected, then the device provides deceleration assistance during the negative power cycle of the user's walking, and at the same time provides acceleration assistance by relying on the elastic potential energy stored in the spring during the positive power cycle, but the power generation power of the device will be reduced under this design parameter. The wearable energy harvesting assistance device can achieve an optimal balance between the energy recovery efficiency and the assistance effect according to the specific requirements of the user, improving the daily user experience of the wearable energy harvesting device. Description of the Drawings

[0046] Figure 1 is a schematic flow chart of a customized multi-objective optimization design method for a wearable energy harvesting assistance device provided by the present invention;

[0047] Figure 2 It is a schematic diagram of the model parameters of the knee joint energy harvesting and assisting device in the embodiments of the present invention;

[0048] Figure 3 It is a schematic diagram of the mechanical structure and the energy harvesting circuit of the knee joint energy harvesting and assisting device in the embodiments of the present invention;

[0049] Figure 4 It is the Pareto front of the multi-objective optimization design of the knee joint energy harvesting and assisting device in the embodiments of the present invention. Detailed implementation manners

[0050] Detailed implementation manner one: Combining Figures 1 to 4 To illustrate this implementation manner, the following steps are included in this implementation manner:

[0051] Step one: Record the weight of the target wearer, the lower limb kinematics and plantar pressure data within a single gait cycle;

[0052] Among them, a single gait cycle of the wearer can be divided into n sampling points; the kinematic data within a single gait cycle are three one-dimensional arrays with a length of n, including the angles θ(n), angular velocities ω(n) and angular accelerations α(n) of each joint of the wearer's lower limb, as well as the linear accelerations Acc(n) of each limb segment of the lower limb, with the units being rad, rad / s, rad / s 2 and m / s 2 respectively;

[0053] n is the length of the array. Since a single gait cycle is divided into n sampling points, these arrays are all one-dimensional arrays containing n data; that is to say, the number of sampling points is the array length;

[0054] Step two: Calculate the joint power data of the target wearer based on the weight, lower limb kinematic data, and plantar pressure data;

[0055] Among them, calculating the joint power data of the target wearer includes the following steps:

[0056] Step 2-1: Consider the single leg of the wearer participating in data collection as a planar serial three-link model composed of three joints, namely the hip, knee, and ankle joints, and three links, namely the thigh, calf, and foot;

[0057] Step 2-2: Using the dynamic calculation method, combining the angles θ(n), angular velocities ω(n), angular accelerations α(n) of each joint, the linear accelerations Acc(n) of each limb segment, the plantar pressure data GRF(n), and the weight m of the wearer, solve the joint torque data T(n) of each joint of the lower limb, with the unit being Nm;

[0058] Step 2-3: Based on the torque data T(n) of each lower limb joint and the angular velocity ω(n) of each lower limb joint, calculate the power P(n) of each lower limb joint, with the unit of W;

[0059] Step 3: According to the configuration of the selected energy-harvesting assistive device, determine the power and assistive output model of the device;

[0060] Step 4: Set the optimization goal for the power generated by electricity and the optimization goal for the assistive effect of the device;

[0061] Step 5: Set the device power constraint, spring torque constraint, and cable tension constraint;

[0062] Step 6: Establish a multi-objective optimization design problem, conduct optimization calculations, and obtain the Pareto front of the optimization design. There are two opposite optimization design directions for the optimal assistive configuration and the optimal training configuration on this Pareto front; among them, the optimal assistive configuration aims to optimize the power generation efficiency while reducing the metabolic level of the wearer, and is suitable for users who need daily movement assistance; the optimal training configuration aims to maximize the power generation efficiency while increasing the metabolic level of the wearer, and is suitable for users with daily physical training needs;

[0063] As described above, the multi-objective optimization design problem is specifically expressed as:

[0064] min f(x)=[f1(x),f2(x)] T

[0065] s.t.g j (x)≤0,j=1,2,3.

[0066] Among them, x is the decision vector for multi-objective optimization.

[0067] Step 7: According to the actual power generation demand or assistive demand of the user, select customized optimal design parameters on the Pareto front;

[0068] Step 8: Adjust the energy-harvesting device according to the selected parameters to obtain the customized optimal assistive effect.

[0069] This embodiment provides a customized multi-objective optimization design method applicable to wearable energy-harvesting assistive devices. Through the multi-objective optimization design for the joint power characteristics of the target user, the wearable energy-harvesting assistive device can, according to the actual needs of the wearer, achieve daily assistance or training effects on the premise of maximizing the energy recovery efficiency, and improve the daily experience of the wearer.

[0070] In step one of this embodiment, the acquisition of kinematic data is not restricted by sensors, and devices such as inertial measurement units or motion capture cameras can be used; the plantar pressure data within a single gait cycle is a one-dimensional array of length n, denoted as GRF(n), with the unit of N; the acquisition of plantar pressure data is not restricted by sensors, and devices such as plantar pressure insoles and six-axis force sensors can be used; the weight data of the wearer is denoted as m, with the unit of kg.

[0071] In view of the individual differences in joint power caused by different body shapes, gait habits, and joint pathological states of different wearers, the present invention proposes a multi-objective optimization design method for customized energy harvesting and assisting devices targeting the joint power characteristics of target customers, so as to customize the assisting characteristics of the energy harvesting and assisting devices according to the needs of the wearers, optimize the assisting effect or power generation efficiency of the devices, and solve the problem that existing energy harvesting devices are difficult to adapt to the physiological characteristic differences among different wearers.

[0072] Specific embodiment two: Combined with Figures 1 to 4 To illustrate this embodiment, the selected configuration of the energy harvesting device in step three of this embodiment is strapped to the outside of the lower limb of the wearer. The energy harvesting device that collects the joint energy of the wearer through a cable-pulley structure and drives a generator to generate electric energy. The radius of the pulley is r, with the unit of m; the energy harvesting device has a spring energy storage mechanism for pre-tightening the cable and providing walking assistance. The elastic coefficient of the spring energy storage mechanism is K, with the unit of Nm / rad.

[0073] With such a setting, the damping force of the generator is relied on to provide deceleration assistance for the negative power cycle movement of the wearer's joint, and the elastic potential energy stored in the spring is relied on to provide acceleration assistance for the positive power cycle movement of the wearer's joint. While realizing full-cycle assistance, the energy recovered during the negative power cycle is converted into electric energy. Other compositions and connection relationships are the same as those in specific embodiment one.

[0074] When determining the power of the device with the knee joint energy harvesting and assisting device as an example in step three of this embodiment, it specifically includes the following steps:

[0075] Step three one: Taking the knee joint energy harvesting and assisting device as an example, it is strapped to the outside of the wearer's thigh. One end of the cable structure is fixed on the pulley, and the other end is fixed to the wearer's calf via an anchor point. Then the cable length L and the cable length change speed dL can be expressed as:

[0076]

[0077] In the formula, h1 and h2 are respectively the distances from the fixed position of the energy harvesting device on the thigh to the knee joint center and the distance from the fixed position of the cable anchor point on the calf to the knee joint center; θ k and ω k are respectively the joint angle and angular velocity of the wearer's knee joint.

[0078] Step 32: The rotational speed ω of the generator driven by the wire-pulley mechanism e and the corresponding back electromotive force E, armature current I, damping torque T d can be expressed as:

[0079] ω e = dL / r

[0080]

[0081] T d = k t I

[0082] where k e and k t are the back electromotive force constant and torque constant of the generator respectively, and R0 and R l are the internal resistance of the generator and the load resistance of the energy harvesting circuit respectively.

[0083] Step 33: The elastic torque T s of the spring energy storage mechanism, the tension F of the wire, and the mechanical power P m of the energy harvesting assist device can be expressed as:

[0084]

[0085] F = (T s + T d ) / r

[0086] P m = FdL

[0087] where θ0 is the pre-tightening angle of the spring energy storage mechanism, and ΔL is the length change of the wire compared to the initial state.

[0088] Step 34: P m is the theoretical calculation data of the mechanical power of the energy harvesting assist device within a single gait cycle, and it should be a one-dimensional array of length n.

[0089] At this time, compare the positive and negative sign relationship of the knee joint power P k (n) and the mechanical power P m of the energy harvesting assist device point by point; when the signs of P k (i) and P m (i) at the i-th sampling point are opposite, it indicates that at this time the energy harvesting assist device is recovering the negative work of the wearer's knee joint or is helping the wearer's knee joint to move by releasing the elastic potential energy of the elastic energy storage mechanism. At this time, the effect of the energy harvesting assist device helps to improve the wearer's exercise effect; when the signs of P k (i) and P mWhen the signs of (i) are the same, it indicates that the energy harvesting assistive device is hindering the movement of the wearer. At this time, the effect of the energy harvesting assistive device will increase the difficulty of the wearer's movement. The above symbol comparison process can be expressed as:

[0090]

[0091] In the formula, P + (n) is the power data of the energy harvesting assistive device hindering the movement of the wearer in a single gait cycle, and P - (n) is the power data of the energy harvesting assistive device assisting the movement of the wearer in a single gait cycle.

[0092] With such a setting, the device power P m can be divided into a positive power P + and a negative power P - two one-dimensional arrays, which is convenient for subsequent operations.

[0093] Specific Embodiment Three: Combining Figures 1 to 4 to illustrate this embodiment, for the electrical energy generation power optimization target f1(x) in step four of this embodiment, it is quantified by the average electrical power of the generator within a single gait cycle, and is expressed as:

[0094]

[0095] In the formula, x is the decision vector of multi-objective optimization, E is the voltage generated by the device, and I is the current generated by the device.

[0096] With such a setting, the theoretical data of the electrical power generated by the device within a single gait cycle can be calculated. The other components and connection relationships are the same as those in Specific Embodiment One or Two.

[0097] Specific Embodiment Four: Combining Figures 1 to 4 to illustrate this embodiment, for the device assistance effect optimization target f2(x) in step four of this embodiment, it is measured by the change in the metabolic level of the wearer after being affected by the energy harvesting assistive device, and is specifically expressed as:

[0098]

[0099] Among them, P + is the power of the positive work done by the energy harvesting device on the human body, and P - is the power of the negative work done by the energy harvesting device on the human body.

[0100] With such a setting, the theoretical value of the change in the metabolic level of the device within a single gait cycle can be calculated. The other components and connection relationships are the same as any one of Specific Embodiments One to Three.

[0101] Specific Embodiment Five: Combining Figures 1 to 4Regarding this embodiment, the device power constraint g1(x) in step five of this embodiment is specifically expressed as:

[0102]

[0103] where P k is the biological power of the human knee joint, and P m is the power exerted by the energy harvesting device on the human body.

[0104] With such a setting, it is ensured that the mechanical energy of the device does not exceed the joint power of the human body itself, ensuring the correctness of the optimal design calculation. The other compositions and connection relationships are the same as any one of the first to fourth specific embodiments.

[0105] Specific embodiment six: Combining Figures 1 to 4 Regarding this embodiment, the spring torque constraint g2(x) and the cable tension constraint g3(x) in step five of this embodiment are specifically expressed as follows to ensure that the cable is always in a pre-tightened state, enabling the energy harvesting assist device to always follow the wearer's movement and maximizing the assist efficiency:

[0106]

[0107] where Ts is the torque of the scroll spring and F is the cable tension.

[0108] With such a setting, it is ensured that the cable is always in a pre-tightened state, enabling the energy harvesting assist device to always follow the wearer's movement and maximizing the assist efficiency. The other compositions and connection relationships are the same as any one of the first to fifth specific embodiments.

[0109] Specific embodiment seven: Combining Figures 1 to 4 Regarding this embodiment, the multi-objective optimization problem in step six of this embodiment is iteratively solved using a multi-objective optimization intelligent algorithm to obtain multiple sets of optimal design parameters that meet the requirements of each optimization objective.

[0110] In the optimal design of the wearable knee joint energy harvesting assist device, the decision vector x mentioned in the multi-objective optimization problem includes: the distance h1 from the fixed position of the energy harvesting assist device on the wearer's thigh to the center of the wearer's knee joint, the distance h2 from the fixed position of the cable anchor point on the wearer's calf to the center of the wearer's knee joint, the radius r of the cable pulley, the load resistance R of the energy harvesting circuit l , the elastic coefficient K of the elastic energy storage mechanism, and the spring pre-tightening angle θ0 of the elastic energy storage mechanism;

[0111] It is specifically expressed as:

[0112] x = [h1, h2, r, R l , K, θ0] T .

[0113] With such settings, the multi-objective optimization problem is iteratively solved by using multi-objective optimization intelligent algorithms such as NSGA-3 and MOPSO to obtain multiple sets of optimized design parameters that meet the requirements of various optimization objectives. The other compositions and connection relationships are the same as any one of the first to seventh specific embodiments.

[0114] Specific embodiment eight: Combining Figures 1 to 4 To illustrate this embodiment, in step seven of this embodiment, according to the actual power generation demand or assistance demand of the user, customized optimal adjustment parameters are selected on the Pareto front, that is, according to the needs of the target user, the optimized design parameters of the wearable energy harvesting assistance device are selected from the optimal assistance configuration and the optimal training configuration. The optimized design parameters include: the distance h1 from the fixed position of the energy harvesting assistance device on the wearer's thigh to the center of the wearer's knee joint, the distance h2 from the fixed position of the cable anchor point on the wearer's calf to the center of the wearer's knee joint, the radius r of the cable pulley, the load resistance R of the energy harvesting circuit l , the elastic coefficient K of the elastic energy storage mechanism, and the spring pre-tightening angle θ0 of the elastic energy storage mechanism.

[0115] In step eight of this embodiment, the energy harvesting device is adjusted according to the selected parameters to obtain a customized optimal assistance effect, that is, according to the optimized design parameters selected in step seven, the design and optimization adjustment of the energy harvesting assistance device are carried out for the target customer to meet the individual differences among different wearers.

[0116] With such settings, the user can further view the optimized design parameters according to the data points selected on the Pareto front, and adjust the wearable energy harvesting device under the guidance of the optimized design parameters to achieve different expected assistance effects. The other compositions and connection relationships are the same as any one of the first to seventh specific embodiments.

[0117] Combining Figures 1 to 4 To illustrate the working principle of the present invention:

[0118] In this embodiment, a single gait cycle of the wearer is divided into 100 sampling points, and the knee joint physiological data collected and calculated through step one and step two are as follows:

[0119] θ k= [0.1182, 0.1306, 0.1440, 0.1600, 0.1800, 0.2048, 0.2326, 0.2602, 0.2838, 0.3003, 0.3089, 0.3107, 0.3078, 0.3025, 0.2966, 0.2908, 0.2852, 0.2792, 0.2720, 0.2631, 0.2523, 0.2397, 0.2258, 0.2111, 0.1961, 0.1813, 0.1667, 0.1525, 0.1387, 0.1256, 0.1133, 0.1018, 0.0915, 0.0824, 0.0747, 0.0683, 0.0633, 0.0597, 0.0575, 0.0566, 0.0574, 0.0598, 0.0641, 0.0705, 0.0792, 0.0904, 0.1041, 0.1203, 0.1392, 0.1607, 0.1852, 0.2132, 0.2455, 0.2830, 0.3267, 0.3769, 0.4335, 0.4954, 0.5608, 0.6273, 0.6927, 0.7548, 0.8124, 0.8648, 0.9119, 0.9535, 0.9895, 1.0197, 1.0436, 1.0611, 1.0720, 1.0764, 1.0747, 1.0672, 1.0541, 1.0358, 1.0124, 0.9840, 0.9505, 0.9119, 0.8682, 0.8193, 0.7652, 0.7063, 0.6428, 0.5755, 0.5050, 0.4324, 0.3590, 0.2863, 0.2166, 0.1523, 0.0966, 0.0525, 0.0226, 0.0084, 0.0097, 0.0242, 0.0475, 0.0738] T

[0120] ω k＝[1.2562,1.3517,1.6011,2.0108,2.4721,2.7930,2.7935,2.4092,1.7304,0.9488,0.2557,-0.2345,-0.5010,-0.5916,-0.5869,-0.5690,-0.5988,-0.7016,-0.8661,-1.0575,-1.2368,-1.3760,-1.4640 -1.5034,-1.5049,-1.4818,-1.4447,-1.3984,-1.3422,-1.2721,-1.1845,-1.0786,-0.9575,-0.8261,-0.6900,-0.5530,-0.4166,-0.2797,-0.1390,0.0103,0.1735,0.3548,0.5568,0.7789,1.0179,1.2678,1.5228,1.7800,2.0431,2.3251,2.6478,3.0369,3.5129,4.0791,4.7120,5.3599,5.9494,6.4021,6.6536,6.6726,6.4705,6.0972,5.6200,5.0965,4.5578,4.0084,3.4388,2.8395,2.2099,1.5603,0.9082,0.2717,-0.3363,-0.9110,-1.4552,-1.9763,-2.4839,-2.9870,-3.4926,-4.0045,-4.5221,-5.0401,-5.5478,-6.0304,-6.4703,-6.8488,-7.1466,-7.3421,-7.4074,-7.3054,-6.9896,-6.4116,-5.5352,-4.3570,-2.9259,-1.3546,0.1858,1.4889,2.3560,206575] T

[0121] P k=[12.5597,-13.2046,-27.6730,-23.0534,-7.8322,22.3471,45.9842,42.7432,30.4237,19.9304,7.3104,-8.7246,-22.0163,-28.8083,-30.7776,-31.7347,-35.0131,-42.3496,-53.3206,-65.7532,-77.0542,-85.3382,-89.8264,-90.7265,-88.8685,-85.2227,-80.4826,-74.9987,-68.8236,-61.9034,-54.3320,-46.3468,-38.3081,-30.5909,-23.4968,-17.2285,-11.8196,-7.1870,-3.2165,0.2140,3.1917,5.7558,7.8985,9.5669,10.6909,11.1754,11.0021,10.1993,8.8104,6.9329,4.5178,1.4247,-2.5114,-7.5225,-13.8437,-21.2993,-29.3581,-37.5921,-44.9748,-50.4283,-53.5248,-54.2463,-53.0722,-50.3952,-46.4926,-41.6085,-35.6556,-28.8214,-21.4763,-14.1722,-7.4603,-1.9215,1.9222,3.7892,3.4777,0.8316,-4.3187,-12.0071,-22.1567,-34.7407,-49.6709-66.5743,-84.9841,-104.2222,-123.0602,-139.3703,-151.2054,-157.2449,-156.3653,-147.5205,-130.8377,-107.5626,-81.4578,-58.8215,-38.6702,-17.5453,2.3150,17.0607,23.7362,22.8364] T

[0122] Combine Figure 4For this embodiment, it is set that the multi-objective optimization method in step six is NSGA-3 and the population size is 200. After the target wearer's physiological data is calculated through steps three to six, a customized Pareto front is obtained, and two optimization design directions, namely the optimal assistance configuration and the optimal training configuration, can be seen on the Pareto front.

[0123] Under the optimal training configuration requirement, the decision vector x obtained from the multi-objective optimization design is x = [0.3875 m, 0.5 m, 0.005 m, 776.6155 Ω, 5.2187e-4 Nm / rad, 55.0108 rad]. T , and the final theoretical optimization target values are f1(x) = -0.3554 W and f2(x) = 0.0319 W.

[0124] Under the optimal assistance configuration requirement, the decision vector x obtained from the multi-objective optimization design is x = [0.2884 m, 0.5 m, 0.005 m, 604.5058 Ω, 2.7890e-4 Nm / rad, 62.83 rad]. T , and the final theoretical optimization target values are f1(x) = -0.3165 W and f2(x) = -0.2447 W.

[0125] Finally, according to the specific requirements of the target customer, the optimal assistance configuration or the optimal training configuration is selected to perform customized design and adjustment on the wearable energy harvesting assistance device.

[0126] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A customized multi-objective optimization design method applicable to wearable energy harvesting and assisting devices, characterized in that: It includes the following steps: Step 1: Record the weight of the target wearer, the lower limb kinematics and plantar pressure data within a single gait cycle; Among them, a single gait cycle of the wearer can be divided into n sampling points; the kinematic data within a single gait cycle are three one-dimensional arrays of length n, and each one-dimensional array includes the angles θ(n), angular velocities ω(n), and angular accelerations α(n) of each joint of the wearer's lower limbs, as well as the linear accelerations Acc(n) of each limb segment of the lower limbs, with the units being rad, rad / s, rad / s 2 and m / s 2 body; Step 2: Calculate the joint power data of the target wearer based on the weight, lower limb kinematic data, and plantar pressure data; Among them, calculating the joint power data of the target wearer includes the following steps: Step 2-1: Consider the single leg of the wearer participating in data collection as a planar serial three-link model composed of three joints (hip, knee, and ankle) and three links (thigh, calf, and foot); Step 2-2: Using the dynamic calculation method, combined with the joint angles θ(n), angular velocities ω(n), angular accelerations α(n), linear accelerations Acc(n) of each limb segment, plantar pressure data GRF(n), and the wearer's weight m, solve the joint torque data T(n) of each lower limb joint, with the unit of Nm; Step 2-3: Based on the joint torque data T(n) of each lower limb joint and the angular velocity ω(n) of each lower limb joint, calculate the joint power P(n) of each lower limb joint, with the unit of W; Step 3: Determine the power and assistive output model of the device according to the selected configuration of the energy harvesting assistive device; Step 4: Set the optimization goals for the power generation of electrical energy and the assistive effect of the device; Step 5: Set the device power constraint, spring torque constraint, and cable tension constraint; Step 6: Establish a multi-objective optimization design problem, conduct optimization calculations, and obtain the Pareto front of the optimization design. There are two opposite optimization design directions for the optimal assistive configuration and the optimal training configuration on this Pareto front; among them, the optimal assistive configuration aims to optimize the power generation efficiency while reducing the wearer's metabolic level, and is suitable for users who need daily movement assistance; the optimal training configuration aims to maximize the power generation efficiency while increasing the wearer's metabolic level, and is suitable for users with daily physical training needs; As described above, the multi-objective optimization design problem is specifically expressed as: min f(x) = [f1(x), f2(x)] T s.t.g j (x) ≤ 0, j = 1, 2, 3. Among them, x is the decision vector for multi-objective optimization. Step 7: Select customized optimal design parameters on the Pareto front according to the actual power generation demand or assistive demand of the user; Step 8: Adjust the energy harvesting device according to the selected parameters to obtain the customized optimal assistive effect.

2. The customized multi-objective optimization design method for a wearable energy harvesting assistive device according to claim 1, characterized in that: The selected configuration of the energy harvesting device in Step 3 is an energy harvesting device that is strapped to the outside of the wearer's lower limb, uses a cable - pulley structure to collect the joint energy of the wearer, and drives a generator to generate electrical energy. The radius of the pulley is r, with the unit of m; the energy harvesting device has a spring energy storage mechanism for pre-tightening the cable and providing movement assistance, and the elastic coefficient of the spring energy storage mechanism is K, with the unit of Nm / rad.

3. A customized multi-objective optimization design method for a wearable energy harvesting assistive device according to claim 2, characterized in that: The optimization goal f1(x) for the power generation of electrical energy in Step 4 is quantified by the average electrical power of the generator within a single gait cycle, and is expressed as: In the formula, x is the decision vector for multi-objective optimization, E is the voltage generated by the device, and I is the current generated by the device.

4. A customized multi-objective optimization design method for a wearable energy harvesting assistive device according to claim 3, characterized in that: The optimization goal f2(x) for the assistive effect of the device in Step 4 is measured by the change in the metabolic level of the wearer after being affected by the energy harvesting assistive device, and is specifically expressed as: Among them, P + is the power of the energy harvesting device doing positive work on the human body, and P - is the power of the energy harvesting device doing negative work on the human body.

5. A customized multi-objective optimization design method for a wearable energy harvesting assistive device according to claim 4, characterized in that: The device power constraint g1(x) in Step 5 is specifically expressed as: Among them, P k is the biological power of the human knee joint, and P m is the power exerted by the energy harvesting device on the human body.

6. A customized multi-objective optimization design method for a wearable energy harvesting assistive device according to claim 5, characterized in that: The spring moment constraint g2(x) and the cable tension constraint g3(x) described in Step 5 are to ensure that the cable is always in a pre-tensioned state, enabling the energy-harvesting assist device to always follow the movement of the wearer and maximizing the assist efficiency. Specifically, they are expressed as: where Ts is the torque of the scroll spring and F is the cable tension.

7. A customized multi-objective optimization design method for wearable energy harvesting assistive devices according to claim 6, characterized in that: The multi-objective optimization problem in Step 6 is iteratively solved using a multi-objective optimization intelligent algorithm to obtain multiple sets of optimized design parameters that meet the requirements of various optimization objectives.

8. A customized multi-objective optimization design method for a wearable energy harvesting assistive device according to claim 7, characterized in that: In step six, in the optimal design of the wearable knee joint energy harvesting and assisting device, the decision vector x in the multi-objective optimization problem includes: the distance h1 from the fixed position of the energy harvesting and assisting device on the wearer's thigh to the center of the wearer's knee joint, the distance h2 from the fixed position of the cable anchor point on the wearer's calf to the center of the wearer's knee joint, the radius r of the cable pulley, and the load resistance R of the energy harvesting circuit l , the elastic coefficient K of the elastic energy storage mechanism, and the spring pre-tightening angle θ0 of the elastic energy storage mechanism; It is specifically expressed as: x = [h1, h2, r, R l , K, θ0] T .

9. A customized multi-objective optimization design method for a wearable energy harvesting assistive device according to claim 1 or 8, characterized in that: In step 7, according to the actual power generation demand or assistance demand of the user, customized optimal adjustment parameters are selected on the Pareto front, that is, according to the needs of the target user, the optimized design parameters of the wearable energy harvesting assistance device are selected from the optimal assistance configuration and the optimal training configuration. The optimized design parameters include: the distance h1 from the fixed position of the energy harvesting assistance device on the wearer's thigh to the center of the wearer's knee joint, the distance h2 from the fixed position of the cable anchor point on the wearer's calf to the center of the wearer's knee joint, the radius r of the cable pulley, and the load resistance R of the energy harvesting circuit l , the elastic coefficient K of the elastic energy storage mechanism, and the spring pre-tightening angle θ0 of the elastic energy storage mechanism.

10. A customized multi-objective optimization design method for a wearable energy harvesting assistive device according to claim 9, characterized in that: In Step 8, the energy-harvesting device is adjusted according to the selected parameters to obtain an optimal customized assist effect, that is, according to the optimized design parameters selected in Step 7, the energy-harvesting assist device is designed and optimized for the target customer to meet the individual differences among different wearers.

Images