A material identification optimization method and system based on RFID radio frequency

Abstract

The present invention discloses a material identification optimization method and system based on RFID radio frequency, which relates to the field of RFID radio frequency identification technology. A material identification optimization system based on RFID radio frequency includes: a material identification optimization module and a special signal processing module. The present invention can adaptively optimize the identification strategy under different identification environments and conditions based on the analysis of radio frequency signal receiving environment data and material identification dynamic data, thereby improving the accuracy of material identification; the material identification equipment can dynamically adjust its working parameters according to the optimization strategy to ensure adaptation to changes in different environments and improve the flexibility and efficiency of identification; through the material identification resource scheduling model, it can intelligently allocate identification tasks and equipment resources. After the material identification omission signal is triggered, a special identification strategy and dynamic adjustment equipment are adopted to ensure that omissions in the identification process are compensated in time.

Description

Technical Field

[0001] The present invention relates to the field of RFID radio frequency identification technology, and in particular to a material identification optimization method and system based on RFID radio frequency. Background Art

[0002] As the demand for refined material management in industries such as logistics, warehousing, and manufacturing continues to increase, RFID technology, due to its non-contact, automated, and real-time characteristics, has become an important means of material identification and tracking. However, in complex application environments, such as large-scale warehousing, outdoor environments, and dense tag deployment, RFID systems often face problems such as multi-tag conflicts, frequency interference, and inaccurate positioning, which affect identification efficiency and accuracy.

[0003] To overcome these problems, multi-band scheduling, positioning technology fusion, dynamic scheduling and adaptive technology can be combined to optimize the RFID material identification process in different environments and application scenarios. Summary of the Invention

[0004] The present invention aims to provide a material identification optimization method and system based on RFID radio frequency, so as to improve the precision and accuracy of RFID radio frequency identification in complex environments.

[0005] A material identification optimization method based on RFID radio frequency includes the following steps:

[0006] Based on the current material recognition scene, radio frequency signal receiving environment data and material recognition dynamic data are obtained; the current material recognition scene contains at least N material recognition devices B n Perform material identification, n=1, 2, ..., N; analyze based on radio frequency signal receiving environment data, material identification dynamic data and material identification environment optimization model to obtain the adaptive material identification optimization strategy L n Among them, L n ={J n , U n , V n}, J n Indicates material identification device B n The shortest material recognition interval, U n Indicates material identification device B n Material recognition power, V n Indicates material identification device B n The material recognition environment optimization model optimizes the working environment of material recognition equipment through environmental data analysis, dynamic prediction and adaptive strategy adjustment, improves recognition accuracy and effectively responds to changing recognition conditions;

[0007] Optimize strategy based on adaptive material identification L nMaterial identification equipment B n Make dynamic adjustments and perform material identification tasks; when a material identification omission signal is received, analyze it based on the material identification resource scheduling model to obtain a special material identification strategy and dynamic material identification equipment; the material identification resource scheduling model ensures that the material identification process is efficient, accurate and without omissions through signal data analysis, equipment matching and dynamic strategy adjustment, thereby optimizing the use of material identification equipment resources and improving identification accuracy; dynamically adjust the dynamic material identification equipment based on the special material identification strategy until a material identification task is completed.

[0008] As a preferred technical solution of the present invention, the material identification environment optimization model includes an environmental data analysis layer, a material identification prediction layer, a dynamic strategy output layer and a strategy output layer;

[0009] The environmental data analysis layer is used to perform feature recognition on the radio frequency signal receiving environment data and obtain the radio frequency signal receiving environment data feature T n ; Among them, the radio frequency signal receiving environment data characteristics T n Indicates the current material identification device B n Corresponding material identification environment characteristic data;

[0010] The material identification prediction layer is used to perform prediction analysis based on material identification dynamic data to obtain the current material identification prediction characteristics;

[0011] The dynamic strategy output layer is used to receive the environmental data characteristics T according to the radio frequency signal. n The dynamic strategy output is carried out with the current material identification prediction characteristics to obtain the adaptive material identification optimization strategy L n ;

[0012] The strategy output layer is used to output the adaptive material recognition optimization strategy L n .

[0013] As a preferred technical solution of the present invention, the specific steps of performing dynamic policy output in the dynamic policy output layer include:

[0014] Build a material identification digital twin model based on the current material identification scenario;

[0015] Construct K material and equipment strategy output individuals G k , k=1, 2, ..., K; each material and equipment strategy outputs individual G k All material identification equipment B n The corresponding simulated material identification optimization strategy; output K material equipment strategies to individual G k Combination, get the material and equipment strategy output iterative population; material and equipment strategy output individual G k The corresponding fitness is Sk ; Fitness S k Output individual G for material and equipment strategy k The corresponding material simulation recognition accuracy; set the maximum number of iterations to M, where m is the current number of iterations, m=1, 2, ..., M;

[0016] The simulation calculation is performed based on the current material recognition prediction features, material recognition digital twin model and swarm optimization algorithm. The specific steps are:

[0017] When the current number of iterations is m, the optimal solution set X of the population is screened. m =[X m1 , X m2 ,…,X mE ], E is the total number of better solutions in the population better solution set, e=1, 2, ..., E; the population better solution set X m The corresponding better fitness set is P m =[P m1 , P m2 ,…,P mE ] T ;

[0018] Using formula W e =(P m avg -P me ) / (P m max -P m min ) calculate the dynamic optimization factor; P m avg Represents the better fitness set P m The average value, P m max Represents the better fitness set P m The maximum value in P m min Represents the better fitness set P m The minimum value in ;

[0019] Calculate the update formula of prey position in population iteration based on dynamic optimization factor R m , R m =W1*X m1 +W2*X m2 +…+W E *X mE ; Use population iteration to update the prey position formula R m Changing the population iteration prey location;

[0020] If P m max -P m min=0, the position of the prey in the population iteration remains unchanged;

[0021] When the maximum number of iterations is reached, the material and equipment strategy output individual corresponding to the current maximum fitness is output, that is, the optimal material and equipment strategy output individual; based on the optimal material and equipment strategy output individual, the adaptive material identification optimization strategy L is obtained. n .

[0022] As a preferred technical solution of the present invention, the specific steps of training the material recognition prediction layer include:

[0023] The material identification and prediction layer includes the data feature splitting layer, the vector splicing layer and the prediction layer;

[0024] The data feature splitting layer is used to split the material identification dynamic data into data features to obtain the material identification dynamic data splitting features;

[0025] The vector splicing layer is used to perform hierarchical division and splicing based on all material identification dynamic data splitting features to obtain a material identification dynamic data splitting feature vector with hierarchical division;

[0026] The prediction layer is used to perform prediction analysis based on the split feature vector of the dynamic data of material identification to obtain the current material identification prediction features;

[0027] Collect several groups of material recognition prediction training samples; each group of material recognition prediction training samples contains current material recognition basic data and corresponding material recognition prediction features; combine several groups of material recognition prediction training samples to obtain a material recognition prediction training set;

[0028] The material identification prediction training set is used for model training to obtain an initial material identification prediction layer; the initial material identification prediction layer is evaluated. If the initial material identification prediction layer passes the model evaluation, the initial material identification prediction layer is used as the material identification prediction layer in the material identification environment optimization model; otherwise, the material identification prediction training set is used to continue model training.

[0029] As a preferred technical solution of the present invention, the material identification resource scheduling model signal data analysis layer, material equipment matching layer, special strategy analysis layer and special strategy output layer;

[0030] The signal data analysis layer is used to obtain the current material trajectory information based on the missing signal of material identification;

[0031] The material equipment matching layer is used to match the material identification equipment according to the current material trajectory information to obtain the dynamic material identification equipment;

[0032] The special strategy analysis layer is used to perform strategy analysis on the dynamic material identification equipment based on the current material trajectory information and the material identification omission signal to obtain the special identification strategy for material identification;

[0033] The special strategy output layer is used to output special identification strategies for material identification.

[0034] As a preferred technical solution of the present invention, the specific steps of training the special strategy analysis layer include:

[0035] Collect several groups of special strategy output training samples; each group of special strategy output training samples contains the target resource scheduling strategy and material identification features; combine several groups of special strategy output training samples to obtain a special strategy output training set; the material identification features contain material trajectory information features and corresponding material identification equipment strategy features;

[0036] Model training is performed based on the special strategy output training set to obtain the initial special strategy analysis layer; a model evaluation is performed on the initial special strategy analysis layer. If the initial special strategy analysis layer passes the model evaluation, the initial special strategy analysis layer is used as the special strategy analysis layer in the material identification resource scheduling model; otherwise, the model training is continued using the special strategy output training set.

[0037] As a preferred technical solution of the present invention, the swarm optimization algorithm is the Pelican optimization algorithm.

[0038] A material identification optimization system based on RFID radio frequency, comprising:

[0039] The material identification optimization module includes a data acquisition unit and a strategy analysis unit; the data acquisition unit is used to obtain radio frequency signal receiving environment data and material identification dynamic data based on the current material identification scene; the current material identification scene contains at least N material identification devices B n Perform material identification, n=1, 2, ..., N; the strategy analysis unit is used to analyze based on the radio frequency signal receiving environment data, material identification dynamic data and material identification environment optimization model to obtain the adaptive material identification optimization strategy L n Among them, L n ={J n , U n , V n}, J n Indicates material identification device B n The shortest material recognition interval, U n Indicates material identification device B n Material recognition power, V n Indicates material identification device B nThe material recognition environment optimization model optimizes the working environment of material recognition equipment through environmental data analysis, dynamic prediction and adaptive strategy adjustment, improves recognition accuracy and effectively responds to changing recognition conditions;

[0040] Special signal processing module, including material identification unit and special strategy analysis unit; material identification unit is used to optimize strategy L based on adaptive material identification n Material identification equipment B n Make dynamic adjustments and perform material identification tasks; when a material identification omission signal is received, analyze it based on the material identification resource scheduling model to obtain a special material identification strategy and dynamic material identification equipment; the special strategy analysis unit is used to deploy the material identification resource scheduling model, and the material identification resource scheduling model ensures that the material identification process is efficient, accurate and without omissions through signal data analysis, equipment matching and dynamic strategy adjustment, thereby optimizing the use of material identification equipment resources and improving identification accuracy; based on the special material identification strategy, the dynamic material identification equipment is dynamically adjusted until a material identification task is completed.

[0041] The present invention has the following advantages:

[0042] 1. The present invention can adaptively optimize the recognition strategy under different recognition environments and conditions through analysis based on radio frequency signal reception environment data and material recognition dynamic data, thereby improving the accuracy of material recognition; the material recognition equipment can dynamically adjust its working parameters according to the optimization strategy to ensure adaptation to changes in different environments and improve the flexibility and efficiency of recognition; through the material recognition resource scheduling model, it can intelligently allocate recognition tasks and equipment resources, reduce invalid resource consumption, and improve the efficiency and accuracy of the overall recognition process. After the material recognition omission signal is triggered, a special recognition strategy and dynamic adjustment equipment are adopted to ensure that omissions in the recognition process are compensated in a timely manner.

[0043] 2. The present invention generates an adaptive material identification optimization strategy through comprehensive analysis of radio frequency signal receiving environment data and material identification dynamic data. It can dynamically adjust the working parameters of the identification equipment according to changes in the actual environment and identification requirements, improve identification accuracy and cope with complex environmental conditions; the material identification environment optimization model can perform feature analysis on radio frequency signal receiving environment data in real time, identify and respond to signal changes in the environment, and ensure that the equipment maintains efficient identification capabilities in a changing environment. The optimized adaptive strategy can reduce the identification interval, adjust the identification power and intensity, and thus improve the working efficiency and identification accuracy of the material identification equipment, especially in complex and dynamic identification scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1This is a structural diagram of an RFID-based material identification optimization system used in an embodiment of the present invention. DETAILED DESCRIPTION

[0045] In order to enable those skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0046] Example 1, a material identification optimization method based on RFID radio frequency, comprising the following steps:

[0047] Based on the current material recognition scene, radio frequency signal receiving environment data and material recognition dynamic data are obtained; the current material recognition scene contains at least N material recognition devices B n Perform material identification, n=1, 2, ..., N; material identification scenarios can include logistics warehousing management, production line material management, airport luggage identification and other practical scenarios; material identification equipment B n It is an RFID radio frequency identification device;

[0048] Based on the analysis of radio frequency signal receiving environment data, material identification dynamic data and material identification environment optimization model, the adaptive material identification optimization strategy L is obtained. n Among them, L n ={J n , U n , V n}, J n Indicates material identification device B n The shortest material recognition interval, U n Indicates material identification device B n Material recognition power, V n Indicates material identification device B n The material recognition environment optimization model optimizes the working environment of material recognition equipment through environmental data analysis, dynamic prediction and adaptive strategy adjustment, improves recognition accuracy and effectively responds to changing recognition conditions;

[0049] The material identification environment optimization model includes an environmental data analysis layer, a material identification prediction layer, a dynamic strategy output layer, and a strategy output layer;

[0050] The environmental data analysis layer is used to perform feature recognition on the radio frequency signal receiving environment data and obtain the radio frequency signal receiving environment data feature T n ; Among them, the radio frequency signal receiving environment data characteristics T n Indicates the current material identification device B n Corresponding material identification environment characteristic data;

[0051] The material identification prediction layer is used to perform prediction analysis based on material identification dynamic data to obtain the current material identification prediction characteristics;

[0052] The dynamic strategy output layer is used to receive the environmental data characteristics T according to the radio frequency signal. n The dynamic strategy output is carried out with the current material identification prediction characteristics to obtain the adaptive material identification optimization strategy L n ;

[0053] The strategy output layer is used to output the adaptive material recognition optimization strategy L n ;

[0054] Through comprehensive analysis of RF signal reception environment data and material identification dynamic data, an adaptive material identification optimization strategy is generated. This strategy can dynamically adjust the operating parameters of the identification device (such as recognition interval, power, and recognition intensity) according to changes in the actual environment and recognition requirements, improving recognition accuracy and coping with complex environmental conditions. The material identification environment optimization model can perform real-time feature analysis of RF signal reception environment data, identify and respond to signal changes in the environment, such as interference and occlusion, to ensure that the device maintains efficient recognition capabilities in changing environments. Through dynamic data analysis at the material identification prediction layer, potential problems in the identification task, such as signal loss or interference between devices, can be predicted in advance. The dynamic strategy output layer adjusts the strategy based on the prediction results to ensure the efficiency and stability of the identification process. The optimized adaptive strategy can reduce the recognition interval and adjust the recognition power and intensity, thereby improving the working efficiency and recognition accuracy of the material identification device, especially in complex and dynamic identification scenarios.

[0055] The specific steps for dynamic policy output in the dynamic policy output layer include:

[0056] Build a material identification digital twin model based on the current material identification scenario;

[0057] Construct K material and equipment strategy output individuals G k , k=1, 2, ..., K; each material and equipment strategy outputs individual G k All material identification equipment B n The corresponding simulated material identification optimization strategy; output K material equipment strategies to individual G k Combination, get the material and equipment strategy output iterative population; material and equipment strategy output individual G k The corresponding fitness is S k ; Fitness S k Output individual G for material and equipment strategy k The corresponding material simulation recognition accuracy; set the maximum number of iterations to M, where m is the current number of iterations, m=1, 2, ..., M;

[0058] The simulation calculation is performed based on the current material recognition prediction features, material recognition digital twin model and swarm optimization algorithm. The specific steps are:

[0059] The swarm optimization algorithm is the Pelican optimization algorithm;

[0060] When the current number of iterations is m, the optimal solution set X of the population is screened. m =[X m1 , X m2 ,…,X mE ], E is the total number of better solutions in the population better solution set, e=1, 2, ..., E; the population better solution set X m The corresponding better fitness set is P m =[P m1 , P m2 ,…,P mE ] T ;

[0061] Using formula W e =(P m avg -P me ) / (P m max -P m min ) calculate the dynamic optimization factor; P m avg Represents the better fitness set P m The average value, P m max Represents the better fitness set P m The maximum value in P m min Represents the better fitness set P m The minimum value in ;

[0062] Calculate the update formula of prey position in population iteration based on dynamic optimization factor R m , R m =W1*X m1 +W2*X m2 +…+W E *X mE ; Use population iteration to update the prey position formula R m Changing the population iteration prey location;

[0063] If P m max -P m min =0, the position of the prey in the population iteration remains unchanged;

[0064] When the maximum number of iterations is reached, the material and equipment strategy output individual corresponding to the current maximum fitness is output, that is, the optimal material and equipment strategy output individual; based on the optimal material and equipment strategy output individual, the adaptive material identification optimization strategy L is obtained. n ;

[0065] By constructing a digital twin model for material identification, the physical and dynamic characteristics of the material identification scenario are digitized, providing a more accurate simulation environment. This model can simulate the material identification process and adjust strategies under different conditions to optimize material identification accuracy and efficiency. By constructing multiple material equipment strategy output individuals and using a swarm optimization algorithm for iterative calculations, it can automatically adjust and select the optimal strategy. The dynamic optimization factor can intelligently adjust the optimization process to ensure that it can adapt to different changes in the material identification environment and always operate in the optimal state. In each iterative process, the optimal solution set is screened, the dynamic optimization factor is calculated, and the position of the prey in the population iteration is updated to accurately optimize the working strategy of the material identification equipment. This fine-grained optimization mechanism can better cope with changing environmental factors and ensure the efficient completion of the identification task. The iterative and dynamic adjustment of the swarm optimization algorithm can avoid ineffective calculations and resource waste, and in-depth optimization is performed only on the strategies most likely to improve identification accuracy, thereby saving computing resources and accelerating the material identification process.

[0066] The specific steps for training the material recognition prediction layer include:

[0067] The material recognition prediction layer includes a data feature splitting layer, a vector splicing layer, and a prediction layer. The material recognition prediction layer can be established using a CNN model.

[0068] The data feature splitting layer is used to split the material identification dynamic data into data features to obtain the material identification dynamic data splitting features;

[0069] The vector splicing layer is used to perform hierarchical division and splicing based on all material identification dynamic data splitting features to obtain a material identification dynamic data splitting feature vector with hierarchical division;

[0070] The prediction layer is used to perform prediction analysis based on the split feature vector of the dynamic data of material identification to obtain the current material identification prediction features;

[0071] Collect several groups of material recognition prediction training samples; each group of material recognition prediction training samples contains current material recognition basic data and corresponding material recognition prediction features; combine several groups of material recognition prediction training samples to obtain a material recognition prediction training set;

[0072] The material identification prediction training set is used for model training to obtain an initial material identification prediction layer; the initial material identification prediction layer is evaluated; if the initial material identification prediction layer passes the model evaluation, the initial material identification prediction layer is used as the material identification prediction layer in the material identification environment optimization model; otherwise, the material identification prediction training set is used to continue model training;

[0073] Through hierarchical processing such as data feature splitting and vector concatenation, the material identification prediction layer can more meticulously analyze and process dynamic data, extract key features, and make efficient predictions. This structured prediction process improves recognition accuracy and stability, especially when processing complex and multi-dimensional data. The data feature splitting layer can deeply explore the potential information in the dynamic material identification data and effectively split it, which enables the prediction layer to utilize richer and more diverse feature data to improve the prediction accuracy of material identification. The vector concatenation layer hierarchically divides and concatenates different split features, fully integrating the feature information of the dynamic material identification data. This hierarchical feature representation enables the model to capture the complex relationships and dependencies between data, thereby improving the reliability of the prediction results.

[0074] Optimize strategy based on adaptive material identification L n Material identification equipment B n Make dynamic adjustments and perform material identification tasks; when receiving a material identification omission signal, analyze it based on the material identification resource scheduling model to obtain a special material identification strategy and dynamic material identification equipment;

[0075] The receiving scenarios or triggering conditions for material identification missed signals may include the following: In complex environments, material tags may experience signal attenuation or complete loss due to obstacles, metal materials, electronic interference, etc., resulting in the device being unable to correctly identify the material; the recognition interval of the material identification device is set too long, which may cause fast-flowing materials to fail to be recognized in time when passing through the identification device, thereby generating missed signals; the frequency or communication protocol of the material identification device and the tag does not match, resulting in the device being unable to recognize the tags of certain materials, thereby generating missed signals;

[0076] The material identification resource scheduling model ensures efficient, accurate, and comprehensive material identification through signal data analysis, equipment matching, and dynamic strategy adjustments, thereby optimizing the use of material identification equipment resources and improving identification accuracy. Dynamic material identification equipment is dynamically adjusted based on special material identification strategies until a material identification task is completed.

[0077] Material identification resource scheduling model signal data analysis layer, material equipment matching layer, special strategy analysis layer and special strategy output layer;

[0078] The signal data analysis layer is used to obtain the current material trajectory information based on the missing signal of material identification;

[0079] The material equipment matching layer is used to match the material identification equipment according to the current material trajectory information to obtain the dynamic material identification equipment;

[0080] The special strategy analysis layer is used to perform strategy analysis on the dynamic material identification device based on the current material trajectory information and the material identification omission signal to obtain the special material identification strategy; the special strategy analysis layer is established based on the BP neural network model;

[0081] The special strategy output layer is used to output special identification strategies for material identification;

[0082] The specific steps for training a special strategy analysis layer include:

[0083] Collect several groups of special strategy output training samples; each group of special strategy output training samples contains the target resource scheduling strategy and material identification features; combine several groups of special strategy output training samples to obtain a special strategy output training set; the material identification features contain material trajectory information features and corresponding material identification equipment strategy features;

[0084] Model training is performed based on the special strategy output training set to obtain an initial special strategy analysis layer. Model evaluation is performed on the initial special strategy analysis layer. If the initial special strategy analysis layer passes the model evaluation, the initial special strategy analysis layer is used as the special strategy analysis layer in the material identification resource scheduling model. Otherwise, model training is continued using the special strategy output training set.

[0085] Through signal data analysis, equipment matching, and dynamic strategy adjustment, the material identification resource scheduling model can effectively respond to missed signals and environmental changes, and optimize the material identification process. This ensures that material identification equipment can perform tasks efficiently and accurately in changing environments, reduces misidentification and omissions, and improves overall work efficiency. Through real-time analysis of material trajectory information and material identification omission signals, the configuration and working strategy of material identification equipment can be dynamically adjusted according to actual needs to ensure that tasks are carried out continuously and are not limited by equipment performance. This real-time dynamic optimization not only improves the reliability of task execution, but also reduces resource waste. The material equipment matching layer performs precise matching based on the real-time acquired material trajectory information to ensure that dynamic material identification equipment can work in the best position and the most suitable environment. Effective equipment matching reduces material identification omissions and ensures identification accuracy and speed. The special strategy analysis layer can formulate personalized identification strategies based on the current material trajectory information and material identification omission signals through in-depth analysis. This strategy is based on historical data training and dynamic adjustment of the current status, and can flexibly respond to different identification environments and task requirements, thereby improving the quality of task completion.

[0086] In actual usage scenarios, such as in large warehouses, materials need to be accurately and efficiently identified and classified to ensure the accuracy of inventory management, material scheduling and cargo sorting. The type, size, storage location and other characteristics of each material need to be predicted and identified in a timely manner. RFID tags can be used to obtain dynamic data of materials in real time, predict the status of materials, and automatically adjust the working parameters of identification equipment based on the identification data of materials to ensure identification accuracy and reduce omissions.

[0087] Example 2, a material identification optimization system based on RFID radio frequency, see Figure 1 As shown, including:

[0088] The material identification optimization module includes a data acquisition unit and a strategy analysis unit; the data acquisition unit is used to obtain radio frequency signal receiving environment data and material identification dynamic data based on the current material identification scene; the current material identification scene contains at least N material identification devices B n Perform material identification, n=1, 2, ..., N; the strategy analysis unit is used to analyze based on the radio frequency signal receiving environment data, material identification dynamic data and material identification environment optimization model to obtain the adaptive material identification optimization strategy L n Among them, L n ={J n , U n , V n}, J n Indicates material identification device B n The shortest material recognition interval, U n Indicates material identification device B n Material recognition power, V n Indicates material identification device B n The material recognition environment optimization model optimizes the working environment of material recognition equipment through environmental data analysis, dynamic prediction and adaptive strategy adjustment, improves recognition accuracy and effectively responds to changing recognition conditions;

[0089] Special signal processing module, including material identification unit and special strategy analysis unit; material identification unit is used to optimize strategy L based on adaptive material identification n Material identification equipment B nMake dynamic adjustments and perform material identification tasks; when a material identification omission signal is received, analyze it based on the material identification resource scheduling model to obtain a special material identification strategy and dynamic material identification equipment; the special strategy analysis unit is used to deploy the material identification resource scheduling model, and the material identification resource scheduling model ensures that the material identification process is efficient, accurate and without omissions through signal data analysis, equipment matching and dynamic strategy adjustment, thereby optimizing the use of material identification equipment resources and improving identification accuracy; based on the special material identification strategy, the dynamic material identification equipment is dynamically adjusted until a material identification task is completed.

[0090] It should be understood that those skilled in the art may make improvements or modifications based on the above description, and all such improvements and modifications shall fall within the scope of protection of the appended claims. Any portion of this specification not described in detail herein is prior art known to those skilled in the art.

Claims

1. A material identification optimization method based on RFID radio frequency, characterized in that: The following steps are involved: Acquire radio frequency signal receiving environment data and material identification dynamic data based on the current material identification scenario; The current material recognition scene contains at least N material recognition devices B n Perform material identification, n=1, 2, ..., N; Based on the analysis of radio frequency signal receiving environment data, material identification dynamic data and material identification environment optimization model, the adaptive material identification optimization strategy L is obtained. n Among them, L n ={J n , U n , V n }, J n Indicates material identification device B n The shortest material recognition interval, U n Indicates material identification device B n Material recognition power, V n Indicates material identification device B n The material recognition environment optimization model optimizes the working environment of material recognition equipment through environmental data analysis, dynamic prediction and adaptive strategy adjustment, improves recognition accuracy and effectively responds to changing recognition conditions; Optimize strategy based on adaptive material identification L n Material identification equipment B n Dynamic adjustments are made and material identification tasks are performed. When a material identification omission signal is received, analysis is performed based on the material identification resource scheduling model to obtain a special material identification strategy and dynamic material identification equipment. The material identification resource scheduling model ensures that the material identification process is efficient, accurate, and complete through signal data analysis, equipment matching, and dynamic strategy adjustment, thereby optimizing the use of material identification equipment resources and improving identification accuracy. Dynamic material identification equipment is dynamically adjusted based on the special material identification strategy until a material identification task is completed. The material identification environment optimization model includes an environmental data analysis layer, a material identification prediction layer, a dynamic strategy output layer, and a strategy output layer; The environmental data analysis layer is used to perform feature recognition on the radio frequency signal receiving environment data and obtain the radio frequency signal receiving environment data feature T n ; Among them, the radio frequency signal receiving environment data characteristics T n Indicates the current material identification device B n Corresponding material identification environment characteristic data; The material identification prediction layer is used to perform prediction analysis based on material identification dynamic data to obtain the current material identification prediction characteristics; The dynamic strategy output layer is used to receive the environmental data characteristics T according to the radio frequency signal. n The dynamic strategy output is carried out with the current material identification prediction characteristics to obtain the adaptive material identification optimization strategy L n ; The strategy output layer is used to output the adaptive material recognition optimization strategy L n .

2. The RFID-based material identification optimization method according to claim 1, characterized in that: The specific steps for dynamic policy output in the dynamic policy output layer include: Build a material identification digital twin model based on the current material identification scenario; Construct K material and equipment strategy output individuals G k , k=1, 2, ..., K; each material and equipment strategy outputs individual G k All material identification equipment B n The corresponding simulated material identification optimization strategy; output K material equipment strategies to individual G k Combination, get the material and equipment strategy output iterative population; material and equipment strategy output individual G k The corresponding fitness is S k ; Fitness S k Output individual G for material and equipment strategy k The corresponding material simulation recognition accuracy; set the maximum number of iterations to M, where m is the current number of iterations, m=1, 2, ..., M; The simulation calculation is performed based on the current material recognition prediction features, material recognition digital twin model and swarm optimization algorithm. The specific steps are: When the current number of iterations is m, the optimal solution set X of the population is screened. m =[X m1 , X m2 ,…,X mE ], E is the total number of better solutions in the population better solution set, e=1, 2, ..., E; the population better solution set X m The corresponding better fitness set is P m =[P m1 , P m2 ,…,P mE ] T ; Using formula W e =(P m avg -P me ) / (P m max -P m min ) calculate the dynamic optimization factor; P m avg Represents the better fitness set P m The average value, P m max Represents the better fitness set P m The maximum value in P m min Represents the better fitness set P m The minimum value in ; Calculate the update formula of prey position in population iteration based on dynamic optimization factor R m , R m =W1*X m1 +W2*X m2 +…+W E *X mE ; Use population iteration to update the prey position formula R m Changing the population iteration prey location; If P m max -P m min =0, the position of the prey in the population iteration remains unchanged; When the maximum number of iterations is reached, the material and equipment strategy output individual corresponding to the current maximum fitness is output, that is, the optimal material and equipment strategy output individual; based on the optimal material and equipment strategy output individual, the adaptive material identification optimization strategy L is obtained. n .

3. The RFID-based material identification optimization method according to claim 2, characterized in that: The specific steps for training the material recognition prediction layer include: The material identification and prediction layer includes the data feature splitting layer, the vector splicing layer and the prediction layer; The data feature splitting layer is used to split the material identification dynamic data into data features to obtain the material identification dynamic data splitting features; The vector splicing layer is used to perform hierarchical division and splicing based on all material identification dynamic data splitting features to obtain a material identification dynamic data splitting feature vector with hierarchical division; The prediction layer is used to perform prediction analysis based on the split feature vector of the dynamic data of material identification to obtain the current material identification prediction features; Collect several groups of material recognition prediction training samples; each group of material recognition prediction training samples contains current material recognition basic data and corresponding material recognition prediction features; combine several groups of material recognition prediction training samples to obtain a material recognition prediction training set; The material identification prediction training set is used for model training to obtain an initial material identification prediction layer; the initial material identification prediction layer is evaluated. If the initial material identification prediction layer passes the model evaluation, the initial material identification prediction layer is used as the material identification prediction layer in the material identification environment optimization model; otherwise, the material identification prediction training set is used to continue model training.

4. The RFID-based material identification optimization method according to claim 3, characterized in that: Material identification resource scheduling model signal data analysis layer, material equipment matching layer, special strategy analysis layer and special strategy output layer; The signal data analysis layer is used to obtain the current material trajectory information based on the missing signal of material identification; The material equipment matching layer is used to match the material identification equipment according to the current material trajectory information to obtain the dynamic material identification equipment; The special strategy analysis layer is used to perform strategy analysis on the dynamic material identification equipment based on the current material trajectory information and the material identification omission signal to obtain the special identification strategy for material identification; The special strategy output layer is used to output special identification strategies for material identification.

5. The RFID-based material identification optimization method according to claim 4, characterized in that: The specific steps for training a special strategy analysis layer include: Collect several groups of special strategy output training samples; each group of special strategy output training samples contains the target resource scheduling strategy and material identification features; combine several groups of special strategy output training samples to obtain a special strategy output training set; the material identification features contain material trajectory information features and corresponding material identification equipment strategy features; Model training is performed based on the special strategy output training set to obtain the initial special strategy analysis layer; a model evaluation is performed on the initial special strategy analysis layer. If the initial special strategy analysis layer passes the model evaluation, the initial special strategy analysis layer is used as the special strategy analysis layer in the material identification resource scheduling model; otherwise, the model training is continued using the special strategy output training set.

6. The RFID-based material identification optimization method according to claim 5, characterized in that: The swarm optimization algorithm is the Pelican optimization algorithm.

7. A material identification optimization system based on RFID radio frequency, characterized in that: The system applies the RFID-based material identification optimization method according to any one of claims 1 to 6, including: The material identification optimization module includes a data acquisition unit and a strategy analysis unit; the data acquisition unit is used to obtain radio frequency signal receiving environment data and material identification dynamic data based on the current material identification scene; the current material identification scene contains at least N material identification devices B n Perform material identification, n=1, 2, ..., N; the strategy analysis unit is used to analyze based on the radio frequency signal receiving environment data, material identification dynamic data and material identification environment optimization model to obtain the adaptive material identification optimization strategy L n Among them, L n ={J n , U n , V n }, J n Indicates material identification device B n The shortest material recognition interval, U n Indicates material identification device B n Material recognition power, V n Indicates material identification device B n The material recognition environment optimization model optimizes the working environment of material recognition equipment through environmental data analysis, dynamic prediction and adaptive strategy adjustment, improves recognition accuracy and effectively responds to changing recognition conditions; Special signal processing module, including material identification unit and special strategy analysis unit; material identification unit is used to optimize strategy L based on adaptive material identification n Material identification equipment B n Make dynamic adjustments and perform material identification tasks; when a material identification omission signal is received, analyze it based on the material identification resource scheduling model to obtain a special material identification strategy and dynamic material identification equipment; the special strategy analysis unit is used to deploy the material identification resource scheduling model, and the material identification resource scheduling model ensures that the material identification process is efficient, accurate and without omissions through signal data analysis, equipment matching and dynamic strategy adjustment, thereby optimizing the use of material identification equipment resources and improving identification accuracy; based on the special material identification strategy, the dynamic material identification equipment is dynamically adjusted until a material identification task is completed.

Images