A lower limb exoskeleton hip joint trajectory tracking control method and system

Abstract

The application discloses a lower limb exoskeleton hip joint trajectory tracking control method and system, and belongs to the technical field of intelligent robot control. By collecting hip joint angle error data, the improved sparrow search algorithm (SSA) with dynamic parameter adjustment is used to optimize the fuzzy PID controller parameters, the Tent chaotic mapping initialization, the error feedback driven finder proportional dynamic adjustment and the convergence factor adaptive mechanism are combined, the population grouping optimization is realized through GPU parallel calculation, and the adaptive t distribution variation and Tent chaotic disturbance strategy are introduced, so that the algorithm convergence speed and global search ability are effectively improved, the problems of fixed traditional PID parameters, fuzzy PID parameter adjustment experience dependence, slow SSA algorithm convergence and large calculation amount are solved, the trajectory tracking precision, response speed and real-time performance are improved, and the method can adapt to complex motion scenes.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent robot control, more particularly to a lower limb exoskeleton hip joint trajectory tracking control method and system based on an adaptive parallel sparrow search algorithm. BACKGROUND

[0002] At present, lower limb exoskeleton robots have become a research hotspot in the fields of rehabilitation medicine and intelligent robots due to their important role in the rehabilitation training of patients with lower limb motor dysfunction. In the aspect of trajectory tracking control, various methods have been proposed in the prior art, such as model predictive control, PID control, fuzzy control, and iterative learning control. Among them, PID control is widely used in industrial scenarios due to its simple structure and ease of implementation. Fuzzy PID control further improves the adaptability of complex systems by introducing fuzzy rules to adaptively adjust parameters. Intelligent optimization algorithms such as sparrow search algorithm (SSA) are also used for parameter optimization, which has made certain progress in improving control accuracy and response speed. In addition, some studies have attempted to solve the challenges of nonlinearity and uncertainty of exoskeleton systems by combining sliding mode control and neural networks.

[0003] However, traditional PID controllers rely on fixed parameters and are difficult to cope with the high dynamics and individual differences of human motion. Although fuzzy PID control can adaptively adjust parameters, the optimization of its quantization factor and proportional factor still relies on artificial experience, which is inefficient and difficult to achieve global optimization. Although the SSA algorithm has global search capability, its fixed parameters (such as population size and iteration number) have slow convergence speed in complex motion scenarios, and the high-dimensional optimization problem has large computational load, which is difficult to meet the real-time control requirements of lower limb exoskeletons. Existing methods are mostly designed for a single motion mode (such as walking on flat ground), and may have problems such as increased tracking error and delayed response in complex scenarios such as stair climbing and slope motion.

[0004] Therefore, how to propose a lower limb exoskeleton hip joint trajectory tracking control method and system to meet the real-time requirements of lower limb exoskeleton robots in high dynamic and multi-degree-of-freedom trajectory tracking is a problem that needs to be solved by those skilled in the art. SUMMARY

[0005] Therefore, the present application provides a lower limb exoskeleton hip joint trajectory tracking control method and system to improve trajectory tracking accuracy, response speed, and real-time performance, and adapt to complex motion scenarios.

[0006] To achieve the above purpose, the present application adopts the following technical solutions:

[0007] On the one hand, the present application proposes a lower limb exoskeleton hip joint trajectory tracking control method, which includes the following steps:

[0008] Collect hip joint angle data, and obtain error data according to the hip joint angle data;

[0009] Set the initial SSA parameters, and dynamically adjust the SSA algorithm parameters according to the current iteration round c and the error data;

[0010] Optimize the position update formula based on the adjusted SSA algorithm parameters, solve by parallel computing, and output the optimal fuzzy PID parameter set;

[0011] Input the optimal fuzzy PID parameter set into the fuzzy PID controller to realize the tracking control of the hip joint trajectory.

[0012] Preferably, setting the initial SSA parameters obtains the population size N, the maximum number of iterations T, the initial proportion of discoverers r begin , the final proportion of discoverers r end , the preset decision factor p, and the chaotic sequence Tent_array d generated by the Tent chaotic mapping initializes the individual position encoding information dimension by dimension;

[0013] The error data includes trajectory error e(t) and trajectory error change rate ec(t);

[0014] The SSA parameters are dynamically adjusted according to the following formula:

[0015]

[0016] s current = 1-r current ;

[0017]

[0018] In the formula, r current is the current discoverer proportion; s current is the current early warning proportion; λ is the error influence coefficient; e max is the maximum allowed error; α is the convergence factor; α max is the maximum value of the convergence factor; α min is the minimum value of the convergence factor; ec th is the error change rate threshold.

[0019] Preferably, the optimized position update formula includes a discoverer position update formula, a follower position update formula and an early warning position update formula;

[0020] The discoverer position update formula is as follows:

[0021]

[0022] In the formula, Xi j c is the position of the ith discoverer in the jth dimension in the cth iteration; LEVY_F is a Levy flight factor; R2 is a random alarm value; ST is a safety threshold; Q is a normally distributed random number; and L is an all-1 matrix.

[0023] The follower position update formula is as follows:

[0024]

[0025] In the formula, Xi j c is the position of the ith follower in the jth dimension in the cth iteration; Xi j c is the position of the ith follower in the jth dimension in the cth iteration; Xi j c is the position of the ith follower in the jth dimension in the cth iteration; Xi j c is the position of the ith follower in the jth dimension in the cth iteration; + Xi j c is the position of the ith follower in the jth dimension in the cth iteration;

[0026] The early-warning position update formula is as follows:

[0027]

[0028] In the formula, Xi j c is the position of the ith follower in the jth dimension in the cth iteration; i Xi j c is the position of the ith follower in the jth dimension in the cth iteration; g Xi j c is the position of the ith follower in the jth dimension in the cth iteration; w Xi j c is the position of the ith follower in the jth dimension in the cth iteration; k is a random number; and ε is a minimum constant.

[0029] Preferably, the calculation formula of the Levy flight factor is as follows:

[0030]

[0031] In the formula, σ and u are both parameters subject to a normal distribution, β is a normally distributed random number with a mean of 0 and a variance of 1, referred to as a step adjustment factor, and Γ(·) is a gamma function.

[0032] Preferably, the calculation formula of the fitness value is as follows:

[0033]

[0034] In the formula, u(t) is the output of the fuzzy PID controller; t s is the adjustment time; e time (t) is an error in a sampling interval, and w1, w2, w3, and w4 are all penalty weights.

[0035] Preferably, the position update formula is optimized based on the adjusted algorithm parameters, parallel computing is adopted for solving, and an optimal fuzzy PID parameter group is output, including:

[0036] Grouping the population into multiple GPUs;

[0037] In each GPU, the position update of the discoverer, follower and early warninger is performed according to the position update formula;

[0038] Taking a random number rand between (0, 1), when the random number rand is less than the preset decision factor p, adaptive t-distributed variation is applied to the sparrow population; when the random number rand is greater than the preset decision factor p, Tent chaotic disturbance is applied to the sparrow population;

[0039] Comparing the individual position information before and after the variation or disturbance processing, if the processed individual position information is better, the processed individual position information is used to replace the original individual;

[0040] Each GPU calculates the individual fitness and returns a local optimum, merges all GPU results, and updates the global optimal solution;

[0041] Judging whether the termination condition is met, if yes, outputting the global optimal solution, taking the PID parameter group corresponding to the global optimal solution as the optimal fuzzy PID parameter group, otherwise, performing a new round of iteration.

[0042] Preferably, adaptive t-distributed variation is applied to the sparrow population, and the process is as follows:

[0043] X variation =X i +X i ·t(c);

[0044] In the formula, X variation is the new individual position information after variation; X i is the position vector of the i-th sparrow individual in the current iteration; t(c) is a t-distributed value with a degree of freedom of the current round number;

[0045] Applying Tent chaotic disturbance to the sparrow population, including:

[0046] Tent_disturbance = LB d +(UB d -LB d )Tent_array d ;

[0047]

[0048] In the formula, X disturbance is the new individual position after chaotic disturbance; LB d, UB d respectively, are the lower and upper limits of the value of the d-th dimensional parameter.

[0049] In another aspect, the application also provides a lower limb exoskeleton hip joint trajectory tracking control system, comprising:

[0050] a data acquisition module, configured to acquire hip joint angle data and obtain error data according to the hip joint angle data;

[0051] a parameter adjustment module, configured to set initial SSA parameters and dynamically adjust the SSA algorithm parameters according to the current iteration round c and the error data;

[0052] an algorithm solving module, configured to optimize a position update formula based on the adjusted SSA algorithm parameters, solve by parallel computing, and output an optimal fuzzy PID parameter group;

[0053] a tracking control module, configured to input the optimal fuzzy PID parameter group into a fuzzy PID controller and realize tracking control of the hip joint trajectory.

[0054] According to the above technical solution, compared with the prior art, the application provides a lower limb exoskeleton hip joint trajectory tracking control method and system. By collecting hip joint angle error data, using an improved sparrow search algorithm (SSA) with dynamic parameter adjustment to optimize the fuzzy PID controller parameters, combining Tent chaotic mapping initialization, finder proportion dynamic adjustment driven by error feedback, and convergence factor adaptive mechanism, realizing population grouping optimization through GPU parallel computing, and introducing adaptive t-distribution variation and Tent chaotic disturbance strategy, the algorithm convergence speed and global search ability are effectively improved, the problems of fixed traditional PID parameters, fuzzy PID parameter adjustment depending on experience, slow SSA algorithm convergence, and large calculation amount are solved, the trajectory tracking precision, response speed, and real-time performance are improved, and the method can adapt to complex motion scenarios. BRIEF DESCRIPTION OF DRAWINGS

[0055] In order to more clearly illustrate the technical solutions in the embodiments of the application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only embodiments of the application, and for those skilled in the art, other drawings can be obtained without creative labor on the basis of the provided drawings.

[0056] Figure 1 a method flowchart provided by the application;

[0057] Figure 2 a parallel computing sparrow search algorithm solving flowchart provided by the application;

[0058] Figure 3The system architecture diagram provided for this invention. Detailed Implementation

[0059] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0060] On the one hand, such as Figure 1 As shown in the figure, an embodiment of the present invention discloses a method for tracking and controlling the hip joint trajectory of a lower limb exoskeleton, comprising the following steps:

[0061] S1. Collect hip joint angle data and obtain error data based on the hip joint angle data.

[0062] Error data include trajectory error e(t) and trajectory error change rate ec(t).

[0063] Real-time acquisition of hip joint angle, calculation of tracking error e(t)=θ ref (t)-θ act (t) and its rate of change

[0064] Error data is the direct basis for controller parameter optimization. The real-time error e(t) can be used to determine the degree of deviation of the current trajectory from the target, and the error change rate ec(t) reflects the dynamic trend of the error. Both of these factors jointly drive the adjustment of SSA parameters and the optimization of fuzzy PID controller parameters.

[0065] e(t) is used to evaluate the current control accuracy and directly affects the calculation of the fitness function; ec(t) is used to judge the dynamic characteristics of the system. When the error changes drastically (ec(t) is large), it is necessary to enhance the global search to quickly correct the deviation.

[0066] S2. Set initial SSA parameters and dynamically adjust the SSA algorithm parameters based on the current iteration round c and error data.

[0067] The initial SSA parameters are set to obtain the population size N, the maximum number of iterations T, and the initial proportion of discoverers r. begin The proportion of discoverers who terminated their work (r) end The default decision factor is p = 0.8, and the chaotic sequence Tent_array generated using Tent chaotic mapping is used. d Initialize individual location encoding information dimension by dimension.

[0068] Chaotic sequence Tent_array d The (d+1)th dimension chaotic sequence value xd+1 As follows:

[0069]

[0070] d is the dimension number of sparrow individual, including 5 parameter dimensions of fuzzy PID: K p , K i , K d , K e , K ec .

[0071] The sequence generated by the Tent chaotic mapping has ergodicity and non-periodicity, can more uniformly cover the search space compared with random initialization, improves the quality of the initial population, and avoids the problem of insufficient population diversity caused by traditional random initialization.

[0072] The SSA parameters are dynamically adjusted according to the following formula:

[0073]

[0074] s current =1-r current ;

[0075]

[0076] In the formula, r current is the current discoverer proportion; s current is the current early warning proportion; λ is the error influence coefficient; e max is the maximum allowed error; α is the convergence factor; α max is the maximum value of the convergence factor; α min is the minimum value of the convergence factor; ec th is the error change rate threshold.

[0077] In the current discoverer proportion adjustment formula, when |e(t)| is large, the term increases, slowing down the discoverer proportion decline speed, retaining more discoverers to perform global search and quickly locate the optimal solution area; when the error is small, the discoverer proportion accelerates the decline, and the early warning proportion automatically increases, focusing on local fine adjustment and improving control accuracy.

[0078] In the convergence factor adjustment formula, ec(t) reflects the error change trend, when ec(t) is large (error fluctuation is severe), the exponential term tends to 1, and α maintains a high value (such as 0.8), maintaining a large search step; when the error tends to be stable (ec(t) is small), α is nonlinearly decreased to α min with iteration, avoiding premature convergence to local optimum.

[0079] The position updating formula is optimized based on the adjusted SSA algorithm parameter, including a discoverer position updating formula, a follower position updating formula and an early-warning position updating formula.

[0080] The discoverer position updating formula is as follows:

[0081]

[0082] In the formula, is the position of the ith discoverer in the jth dimension in the cth iteration; LEVY_F is a Levy flight factor, and the calculation formula is as follows: σ and u are parameters obeying normal distribution, β is a normal distribution random number with a mean of 0 and a variance of 1, referred to as a step adjustment coefficient, Γ(·) is a gamma function; R2 is a random alarm value; ST is a safety threshold; Q is a normal distribution random number; and L is a full 1 matrix.

[0083] The heavy-tailed step is generated through Levy distribution to realize “frequent exploration in short distance + occasional long distance jump” and balance global and local search.

[0084] The follower position updating formula is as follows:

[0085]

[0086] In the formula, is the position of the ith follower in the jth dimension in the cth iteration; is the value of the jth dimension of the global worst discoverer position in the cth iteration; is the value of the jth dimension of the updated optimal position of the discoverer in the cth+1 iteration; is the value of the jth dimension of the global optimal discoverer position in the cth iteration; A + is the pseudo-inverse of the matrix A, and A is a random matrix with elements of ±1.

[0087] The low fitness (i>N / 2) follower performs random diffusion to avoid algorithm precocity; and the high fitness (i≤N / 2) follower develops around the optimal solution of the discoverer to improve convergence speed.

[0088] The early-warning position updating formula is as follows:

[0089]

[0090] In the formula, is the position of the ith early-warning in the jth dimension in the cth iteration; f i is the fitness value of the ith early-warning; f g is the global optimal fitness value; f w is the global worst fitness value; k is a random number; and ε is a minimum constant.

[0091] The early warningers(f i ≠f g ) close to the optimal solution, and the development is strengthened; the early warningers(f i =f g ) in the middle of fitness are randomly moved to avoid the decline of diversity caused by the aggregation of the group.

[0092] The calculation formula of the fitness value is as follows:

[0093]

[0094] In the formula, u(t) is the output of the fuzzy PID controller; t s is the adjustment time; e time (t) is the error in a sampling interval, and w1, w2, w3, and w4 are penalty weights.

[0095] S3. Optimizing the position updating formula based on the adjusted algorithm parameters, solving by parallel computing, and outputting the optimal fuzzy PID parameter group, referring to Figure 2 , including:

[0096] S31. Grouping the population into multiple GPUs.

[0097] The population is divided into subgroups matching the number of GPUs (each GPU processes a group), and the individual position, fitness, and other information of each subgroup are distributed to the corresponding GPU memory through the data transmission interface. The multi-core parallel computing capability of the GPU is used to make each GPU independently execute the position updating, mutation disturbance, and fitness calculation of the discoverers, followers, and early warningers in the subgroup, and realize the parallelization of the population optimization process.

[0098] S32. In each GPU, the position updating of the discoverers, followers, and early warningers is performed according to the position updating formula.

[0099] S33. Taking a random number rand between 0 and 1, when the random number rand is less than the preset decision factor p, adaptive t-distributed mutation is applied to the sparrow population; when the random number rand is greater than the preset decision factor p, Tent chaotic disturbance is applied to the sparrow population.

[0100] Adaptive t-distributed mutation is applied to the sparrow population, and the process is as follows:

[0101] X variation =X i +X i ·t(c);

[0102] In the formula, X variation is the new individual position information after mutation; Xi Xi is the position vector of the i-th Sparrow individual in the current iteration; t(c) is a value subject to t-distribution with degrees of freedom equal to the current round number;

[0103] Applying Tent chaotic disturbance to the Sparrow population, including:

[0104] Tent_disturbance = LB d +(UB d -LB d )Tent_array d ;

[0105]

[0106] In the formula, X disturbance is the new individual position after chaotic disturbance; LB d and UB d are the lower limit and upper limit of the value of the d-th dimensional parameter, respectively.

[0107] S34. Comparing the individual position information before and after mutation or disturbance, if the processed individual position information is better, replace the original individual position information with the processed individual position information;

[0108] S35. Each GPU calculates the individual fitness value and returns the local optimum, merges all GPU results, and updates the global optimal solution;

[0109] Each GPU returns the top K=5 optimal individuals (based on ascending order of fitness), reducing the amount of data transmission.

[0110] Use B-tree structure to merge all local optimal solutions of GPU. After the host processor receives the local optimal solution list returned by each GPU, construct a B-tree according to the fitness value from small to large, use the efficient sorting and deduplication characteristics of B-tree (each node stores multiple key-value pairs, and the number of pointers in non-leaf nodes is dynamically adjusted according to the fitness range), traverse all nodes to filter out the global optimal solution, and finally output the parameter combination with the smallest fitness value as the global optimal solution.

[0111] S36. Determine whether the termination condition is met, if yes, output the global optimal solution, and use the PID parameter set corresponding to the global optimal solution as the optimal fuzzy PID parameter set, otherwise, perform a new round of iteration.

[0112] Termination conditions include:

[0113] Iteration number meets the standard: c≥T.

[0114] Fitness converges: the global optimal fitness changes less than the threshold value for 10 consecutive generations.

[0115] Error meets the standard: the average trajectory error MAE<0.01 rad.

[0116] S4. input the optimal fuzzy PID parameter set into the fuzzy PID controller to realize tracking control of the hip joint trajectory.

[0117] The optimal fuzzy PID parameter set (including quantization factors K e , K ec and proportional factors K p , K i , K d ) is input into the fuzzy PID controller, the controller collects the error e(t) and the error change rate ec(t) of the actual angle of the hip joint and the reference trajectory in real time, and the error e(t) and the error change rate ec(t) are fuzzified into input quantities in the fuzzy domain through K e , K ec , and the PID parameter adjustment amount ΔK p , ΔK i , ΔK d is obtained based on a preset fuzzy rule table, and the real-time control parameters are generated after the basic parameters are superimposed, the control torque u(t) is calculated to drive the joint motor to execute trajectory tracking, and the newly collected angle data is continuously fed back to the controller to form a closed loop of “error calculation-parameter adjustment-control execution”, thereby realizing dynamic and accurate tracking of the hip joint trajectory.

[0118] On the other hand, as shown in Figure 3 , the application also provides a lower limb exoskeleton hip joint trajectory tracking control system, comprising:

[0119] A data acquisition module is configured to collect hip joint angle data and obtain error data according to the hip joint angle data.

[0120] A parameter adjustment module is configured to set initial SSA parameters and dynamically adjust the SSA algorithm parameters according to the current iteration round c and the error data.

[0121] An algorithm solving module is configured to optimize a position update formula based on the adjusted SSA algorithm parameters, solve the formula by parallel computing, and output an optimal fuzzy PID parameter set.

[0122] A tracking control module is configured to input the optimal fuzzy PID parameter set into a fuzzy PID controller to realize tracking control of the hip joint trajectory.

[0123] The embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments, and the same or similar parts of each embodiment can be referred to each other. For the device disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple, and the related parts can be referred to the method part.

[0124] The foregoing description of the disclosed embodiments enables a person skilled in the art to make or use the application. Modifications of these embodiments will occur to persons of skill in the art, and that the appended claims are intended to cover all such modifications that do not depart from the true spirit and scope of the application. Therefore, the application is not limited to the embodiments shown but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method of lower extremity exoskeleton hip trajectory tracking control, the method comprising: The method comprises the following steps: Collecting hip joint angle data, and obtaining error data according to the hip joint angle data; Setting initialization SSA parameters, according to the current iteration round c and error data, dynamically adjusting the SSA algorithm parameters; setting the initialization SSA parameters to obtain the population size N, the maximum iteration number T, the initial proportion of discoverers r begin , the final proportion of discoverers r end , the preset decision factor p , and using the chaotic sequence generated by the Tent chaotic mapping Initialize individual position encoding information dimension by dimension; The error data comprises trajectory error e ( t ) and rate of change of trajectory error The SSA parameters are dynamically adjusted according to the following formula: ( t ) The SSA algorithm parameters are adjusted, and the position update formula is optimized based on the adjusted SSA algorithm parameters; ； ； ； wherein r current is the current discoverer proportion; s current is the current early warning proportion; is the error influence coefficient; e max is the maximum allowed error; is the convergence factor; is the convergence factor maximum value; is the convergence factor minimum value; The optimal fuzzy PID parameter set is output by parallel calculation, and the optimal fuzzy PID parameter set is input into the fuzzy PID controller to realize tracking control of the hip joint trajectory. th is a threshold for the rate of change of error; The optimized position update formula includes a discoverer position update formula, a follower position update formula, and an early warning position update formula; The discoverer position update formula is as follows:

2. The hip trajectory tracking control method of the lower extremity exoskeleton according to claim 1, characterized in that, The follower position update formula is as follows: The early warning position update formula is as follows: ； wherein, is the position of the ith discoverer in the jth dimension for the c+1th iteration; is the position of the ith discoverer in the jth dimension for the cth iteration; LEVY_F is a Levy flight factor; is a random alarm value; ST is a safety threshold; Q is a normally distributed random number; and L is an all-ones matrix. The calculation formula of the Levy flight factor is as follows: ； wherein is the position of the i-th follower in the j-th dimension in the c+1-th iteration; is the position of the i-th follower in the j-th dimension in the c-th iteration; is the value of the j-th dimension of the global worst finder position in the c-th iteration; is the value of the j-th dimension of the updated optimal position of the finder in the c+1-th iteration; is the value of the j-th dimension of the global optimal finder position in the c-th iteration; is the pseudo-inverse of the matrix A, A is a random matrix with elements ±1; The calculation formula of the fitness value is as follows: ； wherein, is the position of the i-th forewarn in the j-th dimension in the c+1-th iteration; is the position of the i-th forewarn in the j-th dimension in the c-th iteration; is the fitness value of the i-th forewarn; is the global optimal fitness value; is the global worst fitness value; k is a random number; is a very small constant.

3. The hip trajectory tracking control method of the lower extremity exoskeleton according to claim 2, characterized in that, The position update formula is optimized based on the adjusted algorithm parameters, and the optimal fuzzy PID parameter set is output by parallel calculation, including: ； wherein , are parameters subject to a normal distribution, is a normally distributed random number with mean 0 and variance 1, called step size adjustment coefficient, is the gamma function.

4. The hip trajectory tracking control method of the lower extremity exoskeleton according to claim 2, characterized in that, Grouping the population into multiple GPUs; ； wherein u t is a fuzzy PID controller output; t s is a tuning time; e time t is an error over a sampling interval, w 1, w 2, w 3, w 4are all penalty weights.​​ 5. The hip trajectory tracking control method of the lower extremity exoskeleton according to claim 2, characterized in that, In each GPU, the positions of the discoverer, the follower, and the early warning are updated according to the position update formula; rand rand taking a random number between (0, 1) rand , when the random number Compare the individual position information before and after mutation or disturbance processing, and replace the original individual with the processed individual if the processed individual is better; is less than the preset decision factor p , applying adaptive t-distribution variation to the sparrow population; when the random number Each GPU calculates the individual fitness and returns the local optimal solution, merges all GPU results, and updates the global optimal solution; is greater than the preset decision factor p , applying Tent chaotic disturbance to the sparrow population; Determine whether the termination condition is met, and if so, output the global optimal solution, and use the PID parameter set corresponding to the global optimal solution as the optimal fuzzy PID parameter set, otherwise, perform a new round of iteration. Apply adaptive t-distributed mutation to the sparrow population, and the process is as follows: t(c) is a t-distributed value with a degree of freedom of the current round number; 6. The hip trajectory tracking control method of the lower extremity exoskeleton according to claim 5, characterized in that, Apply Tent chaotic disturbance to the sparrow population, including: ； In the formula, is the new individual position information after the variation is applied; is the position vector of the i-th sparrow individual in the current iteration; Including: The data acquisition module is configured to collect hip joint angle data and obtain error data according to the hip joint angle data; ； ； In the formula, is the new individual position after chaotic disturbance; , are the lower and upper limits of the value of the dth dimension parameter, respectively.

7. A lower extremity exoskeleton hip trajectory tracking control system for implementing a lower extremity exoskeleton hip trajectory tracking control method as claimed in any one of claims 1-6, characterized by, The algorithm solving module is configured to optimize the position update formula based on the adjusted SSA algorithm parameters, and output the optimal fuzzy PID parameter set by parallel calculation; The tracking control module is configured to input the optimal fuzzy PID parameter set into the fuzzy PID controller to realize tracking control of the hip joint trajectory. The parameter adjustment module is configured to set initial SSA parameters, dynamically adjust the SSA algorithm parameters according to the current iteration round c and error data. ​ ​

Images