Adaptive calibration method of RFID tag antenna based on complex environment recognition

Abstract

The present application relates to the field of antenna calibration technology, and provides an adaptive calibration method for RFID tag antennas based on complex environment recognition. The method includes: collecting warehouse attribute features, constructing a three-dimensional space based on the features, performing RFID signal interference simulation, and clustering to obtain interference areas; for the purpose of meeting signal requirements, calibrating RFID antenna control parameters for multiple areas, determining calibration strategies, and constructing warehouse location calibration tables; reading personnel query paths, matching target interference areas with calibration strategies; reading external interference factors through a monitoring array, analyzing and outputting optimized calibration strategies, and generating optimized parameter calibration schemes; obtaining real-time location information of RFID readers, inputting schemes, outputting real-time calibration strategies, and adaptively calibrating RFID antennas to solve the technical problem that traditional RFID antenna calibration methods cannot quickly adapt to environmental changes, resulting in low efficiency and poor accuracy of RFID antenna calibration.

Description

Technical Field

[0001] The present application relates to the field of antenna technology, specifically to the field of antenna calibration technology, and in particular to an RFID tag antenna adaptive calibration method based on complex environment recognition. Background Art

[0002] With the rapid development of Internet of Things technology, RFID (Radio Frequency Identification) technology, as a contactless automatic identification technology, has been widely used in logistics, warehousing, manufacturing and other fields. RFID uses wireless radio frequency signals to exchange information between tags and readers, and has the advantages of fast recognition speed, large data volume, and strong dynamic recognition capabilities. However, in complex environments, such as areas with dense metal structures and severe electromagnetic interference, the performance of RFID tag antennas is often seriously affected, resulting in reduced recognition accuracy and shortened recognition distance, making it difficult to meet the needs of modern industrial production for efficient and accurate recognition. Traditional RFID tag antenna calibration methods usually rely on fixed calibration parameters and calibration environments, and cannot effectively cope with complex and changeable actual application scenarios. In complex environments such as metal frame storage cabinets, RFID tag antennas may be affected by multiple factors such as metal shielding, electromagnetic interference, and multipath effects, resulting in signal attenuation, reflection, and collision, which in turn affects the recognition effect. Summary of the invention

[0003] This application provides an RFID tag antenna adaptive calibration method based on complex environment recognition, aiming to solve the technical problem that traditional RFID antenna calibration methods cannot quickly adapt to environmental changes, resulting in low RFID antenna calibration efficiency and poor accuracy. By constructing a parameter calibration scheme and dynamically adjusting the calibration strategy based on real-time positioning information, the calibration analysis time is reduced, and the timeliness and accuracy of the adaptive calibration of RFID antennas in complex environments are improved.

[0004] In view of the above problems, the present application provides an RFID tag antenna adaptive calibration method based on complex environment recognition.

[0005] The present application provides an RFID tag antenna adaptive calibration method based on complex environment recognition, the method comprising: collecting attribute characteristics of a target warehouse, constructing a three-dimensional warehouse space through simulation based on the attribute characteristics, simulating RFID signal interference in the three-dimensional warehouse space, clustering to obtain multiple signal interference areas, wherein obstacle interferences in the same signal interference area are considered the same; for the purpose of meeting expected signal requirements, respectively performing control parameter calibration analysis of the RFID antenna for the multiple signal interference areas, determining multiple parameter calibration strategies, and constructing a warehouse position calibration table based on the multiple signal interference areas and the multiple parameter calibration strategies; reading a predetermined query path for a target person, and converting the predetermined query path into a warehouse position calibration table; The warehouse location calibration table is input into the path for matching, and multiple target interference areas and multiple matching parameter calibration strategies are determined; multiple external interference factors of the multiple target interference areas are read through a multi-element monitoring array, and the multiple matching parameter calibration strategies and the multiple external interference factors are input into a predetermined interference calibrator for analysis, and multiple optimization calibration strategies are output, and an optimization parameter calibration scheme is generated in combination with the multiple target interference areas; the real-time location information of the RFID reader is obtained, and the real-time location information is input into the optimization parameter calibration scheme, and the real-time calibration strategy is output to perform parameter adaptive calibration on the RFID antenna, wherein the RFID reader is embedded with a Beidou positioning chip, and the positioning accuracy meets the expected indicators.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0007] The above-mentioned RFID tag antenna adaptive calibration method based on complex environment recognition first collects various attribute characteristics of the target warehouse and constructs a warehouse three-dimensional space based on these characteristics. In this warehouse three-dimensional space, the propagation of RFID signals is simulated and the areas where the signals may be interfered are identified. These interference areas are clustered into multiple groups, and the interference conditions within each group are considered similar. Subsequently, the RFID antenna control parameter analysis is performed on each signal interference area to find the best parameter setting that can reduce or eliminate these interferences, formulate a parameter calibration strategy for each interference area, and form a warehouse location calibration table, which associates different areas with corresponding calibration strategies. When the target personnel need to query or operate items in the warehouse and set a predetermined query path, the warehouse location calibration table will be matched with this predetermined query path to find the signal interference areas and corresponding calibration strategies that may be encountered on the path. In order to further improve the accuracy of the calibration, the external factors of these interference areas are monitored in real time through a multivariate monitoring array, and this information is input into an interference calibrator together with the previously matched calibration strategy for analysis. This analysis process will output a more optimized calibration strategy to form a final optimized parameter calibration solution. When the RFID reader moves in the warehouse, it will obtain its location information in real time, and select the corresponding calibration strategy from the optimized parameter calibration scheme based on this location information to perform adaptive calibration on the RFID antenna, thereby ensuring that the RFID signal can maintain the best state at any location. This process greatly improves the efficiency and accuracy of warehouse management.

[0008] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0010] Figure 1 It is a flowchart of an RFID tag antenna adaptive calibration method based on complex environment recognition in one embodiment;

[0011] Figure 2 The present invention is a schematic diagram of a flow chart of obtaining multiple signal interference areas of an RFID tag antenna adaptive calibration method based on complex environment recognition in an embodiment. DETAILED DESCRIPTION

[0012] The embodiment of the present application provides an RFID tag antenna adaptive calibration method based on complex environment recognition, thereby solving the technical problem that the traditional RFID antenna calibration method cannot quickly adapt to environmental changes, resulting in low RFID antenna calibration efficiency and poor accuracy.

[0013] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0014] It should be noted that the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules that are not explicitly listed or inherent to these processes, methods, products or devices.

[0015] Examples, such as Figure 1 As shown, the present application provides an RFID tag antenna adaptive calibration method based on complex environment recognition, the method comprising:

[0016] The attribute characteristics of the target warehouse are collected, and a three-dimensional warehouse space is constructed based on the attribute characteristics. RFID signal interference simulation is performed in the three-dimensional warehouse space, and multiple signal interference areas are obtained by clustering, wherein obstacle interferences in the same signal interference area are regarded as the same.

[0017] RFID (Radio Frequency Identification) is a non-contact automatic identification technology. It communicates through radio waves and can automatically identify the target object and obtain relevant data without the need to establish mechanical or optical contact between the identification system and the specific target. RFID technology has the advantages of non-contact, fast reading speed, no wear, no environmental impact, and long life. It consists of a tag (radio frequency card or electronic tag), a reader (reader-writer or interrogator), and an antenna.

[0018] In an embodiment of the present application, the system terminal first collects various physical characteristics of the target warehouse, such as the size, layout, materials, etc. of the warehouse, and uses these physical characteristics as the attribute characteristics of the target warehouse for the simulation construction of the warehouse three-dimensional space. After obtaining the attribute characteristics of the target warehouse, the system terminal uses simulation software to build a highly restored three-dimensional warehouse model as the warehouse three-dimensional space based on these attribute characteristics. This warehouse three-dimensional space will present the internal structure of the warehouse, the distribution of items and other factors in detail. Subsequently, in the constructed warehouse three-dimensional space, the system terminal performs multiple RFID signal propagation and interference simulations, and averages the results of the multiple simulations to obtain multiple signal quality evaluation results, and then performs cluster analysis on these signal quality evaluation results, and obtains multiple signal interference areas by comparing multiple deviation thresholds. In the same signal interference area, the interference effects of all obstacles are regarded as the same, which helps to simplify the problem and concentrate on optimization measures, so as to formulate targeted optimization strategies to improve the overall performance and stability of the system.

[0019] Furthermore, the present application provides attribute characteristics of the target warehouse, including:

[0020] The attribute characteristics of the target warehouse include building structure characteristics, building layout characteristics, building size characteristics, building material characteristics and cargo storage characteristics.

[0021] Preferably, the attribute characteristics of the target warehouse collected by the system terminal include building structure characteristics, building layout characteristics, building size characteristics, building material characteristics, cargo storage characteristics, etc. Among them, the building structure characteristics include the roof structure of the warehouse (such as flat roof, sloping roof, etc.), wall thickness, column position and layout, etc. These characteristics determine the path and attenuation of the RFID signal propagation inside the warehouse. For example, thick concrete walls may significantly block RFID signals, while columns may become reflection points or shadow areas of the signal. Building layout characteristics refer to the layout of shelves, passages, work areas, etc. in the warehouse. Different layouts will directly affect the coverage and signal quality of the RFID signal. For example, a dense shelf layout may cause the signal to be reflected multiple times between goods, increasing signal interference; while wide passages are conducive to signal propagation and reducing blind spots. Building size characteristics include the length, width, height and overall space size of the warehouse. These size characteristics determine the propagation distance and attenuation rate of the RFID signal in the warehouse. Large warehouses require stronger RFID signal transmission power or a denser reader layout to ensure that the signal covers the entire area. Building material characteristics refer to the building materials used in the warehouse. Different building materials in the warehouse (such as metal materials, concrete, wood, etc.) have different degrees of interference with RFID signals. Metal materials may absorb or reflect RFID signals due to their conductivity, causing signal attenuation or interference; while non-conductive materials such as concrete may have a certain blocking effect on the signal, but the impact is relatively small. Understanding these material characteristics helps optimize the layout and parameter settings of RFID. Cargo storage characteristics cover the type, size, density and stacking method of goods stored in the warehouse. The attribute characteristics of goods characterize the characteristics of obstacles in the warehouse. Different types and sizes of goods have different blocking and reflection effects on RFID signals, which in turn affects the coverage and recognition rate of the signal. For example, large metal goods may become a strong source of signal interference, while small non-conductive goods have less impact on the signal.

[0022] Further, if Figure 2 As shown, the present application provides a method for obtaining multiple signal interference areas, including:

[0023] The three-dimensional space of the warehouse is equally divided according to a predetermined grid size to determine a plurality of warehouse grid spaces; multiple RFID signal interference simulations are performed in the plurality of warehouse grid spaces, and a plurality of signal quality evaluation results are obtained after averaging, wherein the signal quality evaluation results include at least signal strength, reading range and reading rate; based on the plurality of signal quality evaluation results, a cluster analysis is performed on the plurality of warehouse grid spaces to determine the plurality of signal interference areas.

[0024] Optionally, the system terminal divides the three-dimensional space of the warehouse into equal parts according to a predetermined grid size to form multiple uniform warehouse grid spaces. This step is the basis for subsequent analysis, which helps to simplify the complex warehouse environment into manageable grid units, so as to facilitate accurate analysis of the signal quality of each area. Among them, the predetermined grid size is set according to the physical size of the warehouse, the performance requirements and accuracy requirements of the RFID, and is a commonly used size. In each divided warehouse grid space, the system terminal performs multiple RFID signal interference simulations based on the simulation software. The number of simulations can be flexibly set according to actual needs to ensure the stability and reliability of the results. After each simulation, the signal quality data in each grid space is recorded, including signal strength, reading range and reading rate, and according to the actual accuracy requirements, more comprehensive evaluation dimensions such as signal-to-noise ratio, bit error rate, sensitivity, and frequency response are selected for recording. Subsequently, the multiple simulation data in each grid space are processed, and the mean of each evaluation index is calculated to eliminate the accidental errors of a single simulation and improve the accuracy of the results. Then the signal quality evaluation results of all grid spaces are integrated to obtain multiple signal quality evaluation results. Afterwards, based on these signal quality evaluation results, the system terminal performs cluster analysis on these signal quality evaluation results by comparing the deviation value of each indicator of two adjacent grid spaces with the corresponding predetermined threshold value, and merges the corresponding grid spaces in each cluster to obtain multiple signal interference areas. These areas represent the differential distribution of signal quality in the warehouse, providing an important reference for subsequent RFID antenna optimization.

[0025] Furthermore, the present application provides cluster analysis of the plurality of warehouse grid spaces, including:

[0026] A first grid space and a second grid space are randomly selected from the multiple warehouse grid spaces, wherein the first grid space and the second grid space are any two adjacent grid spaces from the multiple warehouse grid spaces; a first signal strength, a first reading range and a first reading rate of the first grid space, and a second signal strength, a second reading range and a second reading rate of the second grid space are obtained; if a deviation value between the first signal strength and the second signal strength is less than a predetermined strength threshold, a deviation between the first reading range and the second reading range is calculated, and if the range deviation value is less than a predetermined range threshold, a deviation between the first reading rate and the second reading rate is calculated to obtain a reading rate deviation value; if the reading rate deviation value is less than a predetermined reading rate threshold, the first grid space and the second grid space are merged to obtain a first merged space, wherein a signal quality evaluation result of the first merged space is an average of the signal quality evaluation results of the first grid space and the second grid space; iterative merging is continuously performed until there is no adjacent grid space that meets the merging condition, and the multiple signal interference areas are output based on the merging result.

[0027] Optionally, the system terminal randomly selects two adjacent grid spaces from multiple warehouse grid spaces as the first grid space and the second grid space, and obtains the first signal strength, the first reading range, the first reading rate of the first grid space, and the second signal strength, the second reading range and the second reading rate of the second grid space. Subsequently, the differences between the two grid spaces in the three key indicators of signal strength, reading range and reading rate are compared by calculating the absolute difference. If the deviation of the signal strength of the two grid spaces is lower than the predetermined strength threshold, the system terminal determines their deviation in the reading range. If the deviation of the reading range is also lower than the predetermined range threshold, the deviation of the reading rate will be determined. Only when the deviation of the reading rate is also less than the predetermined reading rate threshold, the system terminal determines that the two grid spaces have similar signal quality and merges the two adjacent grid spaces into a new space, namely the first merged space. When there are evaluation indicators of other dimensions, the system terminal performs the same operation until the deviation of the last evaluation indicator and the corresponding predetermined threshold are compared. The first merged space obtained by merging will inherit the mean of the signal quality evaluation results of the original two grid spaces as its own signal quality evaluation result. This process is repeated over and over again, with each iteration checking and merging adjacent grid spaces that meet the conditions until there are no more grid spaces to merge. Afterwards, based on the results of the iterative merging, the system terminal outputs multiple signal interference areas. These areas are composed of grid spaces with similar signal quality that do not meet further merging conditions. In this way, the system terminal can more accurately identify which areas in the warehouse may have signal interference problems, thereby providing targeted guidance for subsequent RFID antenna optimization.

[0028] For the purpose of meeting expected signal requirements, control parameter calibration analysis of the RFID antenna is performed on the multiple signal interference areas respectively, multiple parameter calibration strategies are determined, and a warehouse location calibration table is constructed based on the multiple signal interference areas and the multiple parameter calibration strategies.

[0029] In one embodiment, in order to eliminate the interference of obstacles in the warehouse to the RFID signal, improve the accuracy and stability of signal reception, and meet the expected signal requirements, the system terminal performs a control parameter calibration analysis of the RFID antenna for each determined signal interference area. This includes retrieving multiple groups of sample calibration data based on big data, extracting sample calibration data that meet the calibration requirements required to achieve the expected signal requirements from these sample calibration data groups to form new multiple groups of sample calibration data, and extracting the most frequently occurring sample calibration data from each group of sample calibration data, and adding them to multiple parameter calibration strategies. Subsequently, multiple signal interference areas and their corresponding parameter calibration strategies are integrated into a warehouse location calibration table. This data table records in detail the specific location, interference type, parameter calibration strategy and other information of each area, which is used for subsequent rapid matching of the target person's predetermined query path.

[0030] Further, the present application provides a calibration strategy for determining multiple parameters, including:

[0031] A first signal interference area is randomly selected from the multiple signal interference areas, and a mean value of a first signal quality evaluation result of the first signal interference area is obtained; based on the expected signal requirement, a deviation analysis is performed on the mean value of the first signal quality evaluation result to determine a first signal quality deviation; with the RFID device attributes and the first signal quality deviation as constraints and with the control parameter calibration of the RFID antenna as the purpose, multiple sample calibration data are acquired based on big data retrieval, wherein the sample calibration data includes sample calibration parameters and sample calibration results; sample calibration parameters corresponding to the sample calibration results that meet the expected calibration requirements are selected to construct a standard calibration parameter set, and the mode of the standard calibration parameter set is selected as a first parameter calibration strategy, which is added to the multiple parameter calibration strategies, wherein the expected calibration requirements include calibration speed, calibration accuracy and calibration stability.

[0032] Preferably, the system terminal randomly selects a signal interference area from the obtained multiple signal interference areas as the first signal interference area, and obtains the mean value of the first signal quality evaluation result of the interference area. Subsequently, based on the expected signal demand, the deviation analysis is performed on the mean value of the first signal quality evaluation result, that is, the difference calculation is performed between each evaluation index in the mean value of the first signal quality evaluation result and each expected evaluation index in the expected signal demand, so as to determine the degree of deviation of the current signal quality relative to the expected demand, that is, the first signal quality deviation. In order to improve this deviation, the system terminal uses the inherent attributes of the RFID device (such as device model, power range, etc.) and the first signal quality deviation as dual constraints, and uses big data analysis technology to retrieve and screen relevant sample calibration data. These sample calibration data contain a variety of possible calibration parameters (such as antenna power, antenna frequency, antenna angle, etc.) and their corresponding calibration effects (i.e., sample calibration results), providing a rich reference basis for the control parameter calibration of the RFID antenna. Among the many sample calibration data, the system terminal screens out sample calibration results that meet the expected calibration requirements (such as faster calibration speed, higher calibration accuracy and stable calibration effect). Based on these sample calibration results that meet the conditions, the system terminal constructs a standard calibration parameter set. The parameters in this set are considered to be the best choice that can effectively improve the performance of the RFID antenna. Afterwards, in order to simplify the calibration process and ensure the universality of the calibration effect, the mode, that is, the parameter value with the highest frequency, is selected from the standard calibration parameter set as the first parameter calibration strategy, and it is added to multiple parameter calibration strategies. In this way, when the RFID antenna needs to be calibrated, the control parameters of the antenna, including antenna power, antenna frequency, and antenna angle, can be quickly and accurately adjusted according to the first parameter calibration strategy to achieve the expected signal quality requirements.

[0033] The predetermined query path of the target person is read, the predetermined query path is input into the warehouse location calibration table for matching, and a plurality of target interference areas and a plurality of matching parameter calibration strategies are determined.

[0034] In one embodiment, when the target person needs to search in the target warehouse, the system terminal will read the target person's predetermined query path. This path is set in advance to guide the target person to efficiently find and obtain the required items in the warehouse. Subsequently, this predetermined query path is input into the constructed warehouse location calibration table. By matching the location in the predetermined query path with the location information in the warehouse location calibration table, the system terminal identifies multiple target interference areas that may be encountered on the predetermined query path, as well as the parameter calibration strategy corresponding to each target interference area. By matching multiple target interference areas and multiple matching parameter calibration strategies, it can be ensured that the target person can smoothly obtain information on the query path, and optimize the query efficiency and experience of the target person in the warehouse.

[0035] The multiple external interference factors of the multiple target interference areas are read by a multi-element monitoring array, the multiple matching parameter calibration strategies and the multiple external interference factors are input into a predetermined interference calibrator for analysis, multiple optimization calibration strategies are output, and an optimization parameter calibration scheme is generated in combination with the multiple target interference areas.

[0036] In one embodiment, the system terminal collects multiple external interference factors of each target interference area through a multi-element monitoring array. These external interference factors include signal strength fluctuations, electromagnetic interference, human activity density, cargo stacking status and other factors, which will affect the efficiency and accuracy of the target personnel when retrieving materials. Subsequently, the determined multiple matching parameter calibration strategies and the collected external interference factors are input into a predetermined interference calibrator for analysis. This predetermined interference calibrator is pre-trained according to the BP neural network. It can output multiple optimization calibration strategies based on the input interference factors and the original calibration strategy, taking into account various factors. These optimization strategies are for specific interference areas and specific interference factors, and are intended to minimize interference and improve signal quality and data transmission efficiency by adjusting the relevant parameters of the RFID antenna. Afterwards, a complete optimization parameter calibration scheme is generated based on these optimization calibration strategies and the specific location, size and other information of each target interference area. This scheme will provide detailed guidance on adjusting the parameters of the RFID antenna to ensure that the target personnel can complete the task smoothly and efficiently when retrieving materials.

[0037] Further, the present application provides a method for constructing a multivariate monitoring array, comprising:

[0038] External interference indicators are configured, wherein the external interference indicators include at least electromagnetic interference, environmental interference and multipath effect; based on the external interference indicators, multiple sensor groups are configured for the multiple target interference areas, and the multi-element monitoring array is integrated and constructed.

[0039] Preferably, when constructing a multi-element monitoring array, the system terminal first identifies which external factors may have an adverse effect on the monitoring target, and configures corresponding indicators according to these external factors, namely, external interference indicators. These external interference indicators include electromagnetic interference, environmental interference, and multipath effects. Among them, electromagnetic interference refers to electromagnetic interference from external electronic devices (such as forklift radios, handheld scanning devices, other wireless communication devices, etc.) and electromagnetic sources outside the warehouse. These interferences can affect the reading distance and accuracy of RFID tags. Environmental interference covers temperature changes, humidity fluctuations, cargo stacking density, and warehouse structures (such as metal shelves, concrete walls) in the warehouse. These environmental factors may cause attenuation or reflection of wireless signals. Multipath effect refers to the reflection, refraction, and scattering of shelves, cargo, and other obstacles during the propagation of wireless signals, which causes the signal to reach the receiving end along multiple paths, thereby causing signal superposition, phase shift, and intensity changes, thereby affecting the reading effect of RFID tags. In order to accurately monitor these external interferences, the system terminal configures multiple sensor groups in multiple target interference areas. These sensor groups include sensors for monitoring electromagnetic field strength, temperature and humidity sensors, and RFID test equipment that can evaluate signal quality. By integrating the data from these sensor groups, a multi-element monitoring array is built, which can monitor external interference in the warehouse in real time and provide data support for subsequent parameter optimization.

[0040] Furthermore, the present application provides a method for constructing a predetermined interference calibrator, comprising:

[0041] Taking the external interference index as a constraint, retrieve and obtain a sample calibration data set, perform differential screening on the sample calibration data set, and obtain a sample training data set, wherein the sample calibration data includes a sample external interference factor, a sample calibration strategy, and a sample optimization calibration strategy; divide the sample training data set into Q equal parts, and select Q times with replacement to construct a first training set, and iterate and select Q times to obtain Q training sets; use the Q training sets to perform supervised training and cross-validation on the BP neural network to obtain Q convergence interference calibration units, and construct the predetermined interference calibrator based on the Q convergence interference calibration units, wherein the output of the predetermined interference calibrator is the mode of the output results of the Q convergence interference calibration units.

[0042] Optionally, the system terminal collects sample calibration data according to the external interference index. These data not only include external interference factors, but also record the calibration strategies taken under different interference conditions and the optimized calibration strategies. Then, these sample calibration data are differentially screened to remove similar data and retain samples that can represent different interference scenarios and calibration effects to form a sample training data set. Subsequently, the screened sample training data set is divided into Q parts, where Q is a preset value representing the number of training sets. Then, a sampling method with replacement is adopted to randomly select Q times from the Q parts of data, and each time a piece of data is selected as part of the training set to construct the first training set. Repeat this process Q times, but the data selected each time may be different, and finally Q different training sets are obtained. After that, the BP neural network is supervised and trained using these Q training sets. The BP neural network is a commonly used feedforward neural network that optimizes the network weights through the back propagation algorithm to achieve the desired output results. During the training process, cross-validation is also performed, that is, each training set is used as a test set to evaluate the performance of the model trained by other training sets to ensure the generalization ability of the model. After training, the system terminal obtains Q converged interference calibration units, each of which is a BP neural network model trained based on different training sets and can be calibrated for specific interference scenarios. Then, the Q converged interference calibration units are connected in parallel to construct a predetermined interference calibrator. When actually used, the predetermined interference calibrator will receive the current external interference factor and matching parameter calibration strategy as input, and process them through Q calibration units respectively to obtain Q output results. In order to improve the accuracy and stability of the calibration, the predetermined interference calibrator uses the majority of the Q output results as the final output, that is, the value that appears the most times in the Q output results is selected as the calibration result. This can reduce the errors caused by individual calibration units due to training data deviation or model overfitting.

[0043] Furthermore, the present application provides a method for obtaining a sample training data set, including:

[0044] A first sample calibration data is randomly selected from the sample calibration data set, wherein the first sample calibration data includes a first sample external interference factor and a first sample calibration strategy; a predetermined similarity comparison algorithm is used to perform a similarity comparison on the first sample external interference factor and other sample external interference factors in the sample calibration data set, and a number whose statistical similarity is less than a predetermined factor similarity threshold is set as a first difference; if the first difference is greater than a predetermined threshold, a similarity comparison is performed on the first sample calibration strategy and other sample calibration strategies in the sample calibration data set, and a number whose statistical similarity is less than a predetermined strategy similarity threshold is set as a second difference; if the second difference is greater than a predetermined threshold, the first sample calibration data is added to the sample training data set.

[0045] Optionally, the system terminal randomly selects a sample calibration data from the sample calibration data set as the first sample calibration data, which includes the first sample external interference factor and the corresponding first sample calibration strategy. This step ensures the randomness of data selection and lays the foundation for subsequent differentiated screening. Subsequently, the first sample external interference factor is used to calculate the similarity with the external interference factors of other samples in the data set using a predetermined similarity comparison algorithm, that is, the distance between the first sample external interference factor and the external interference factors of other samples is calculated by Euclidean distance, and the inverse calculation is performed to obtain multiple similarities. After that, the number of similarities less than the predetermined factor similarity threshold is counted and recorded as the first difference. This predetermined factor similarity threshold is used to determine whether the external interference factor of the first sample is unique enough to distinguish it from other samples in the data set. If the first difference is greater than the predetermined threshold, it means that the external interference factor of the first sample has sufficient difference, and the corresponding first sample calibration strategy can be compared. The system terminal uses the same method to compare the first sample calibration strategy with the calibration strategies of other samples in the data set, and the number of similarities less than the predetermined strategy similarity threshold is counted and recorded as the second difference. This step is to confirm that the calibration strategy of the first sample is also unique and valuable, so as to avoid redundant or repeated calibration strategies from being included in the training data set. If the second difference is also greater than the predetermined threshold, it means that the calibration data of the first sample has sufficient differences in both external interference factors and calibration strategies, so it is added to the sample training data set. Through such a differentiated screening process, it can be ensured that the training data set contains more representative and unique samples, thereby improving the efficiency and accuracy of training. The whole process is to perform a similarity dimensionality reduction process on the retrieved data, that is, to remove samples with high similarity and lack of uniqueness through similarity comparison and differentiated screening. This process reduces the redundancy and repetition in the training data, so that useful features and information can be learned faster during the training process, thereby reducing training time and improving training efficiency. At the same time, since the training data set is more refined and representative, the training accuracy will also be improved.

[0046] The real-time location information of the RFID reader is obtained, the real-time location information is input into the optimized parameter calibration scheme, and a real-time calibration strategy is output to perform parameter adaptive calibration on the RFID antenna, wherein the RFID reader is embedded with a Beidou positioning chip, and the positioning accuracy meets the expected indicators.

[0047] In one embodiment, in order to achieve accurate and real-time calibration of RFID antenna parameters, the system terminal uses an RFID reader with an integrated Beidou positioning chip. This Beidou positioning chip can provide accurate location information of the reader in real time, and its positioning accuracy reaches the set expected standard. The system terminal continuously obtains and updates the current location information of the reader through the Beidou positioning chip built into the RFID reader, and inputs this real-time location information into the obtained optimization parameter calibration scheme. The system terminal automatically adjusts and optimizes the relevant parameters of the RFID antenna according to the optimization parameter calibration scheme and the position change of the reader to adapt to different environmental conditions and position changes. Through such a real-time calibration strategy, the RFID antenna can always maintain the best working state to ensure accurate reading and identification of RFID tags. This adaptive calibration method not only improves the overall performance of RFID, but also reduces the recognition error caused by environmental changes, so that RFID technology can play the best role in various application scenarios.

[0048] In summary, the embodiments of the present application have at least the following technical effects:

[0049] The embodiment of the present application first collects the attribute characteristics of the target warehouse, constructs a three-dimensional simulation space based on these characteristics, and performs RFID signal interference simulation therein, obtains multiple signal interference areas through cluster analysis, and analyzes the obstacle interference in these areas. Subsequently, for each signal interference area, the control parameter calibration analysis of the RFID antenna is performed, and a warehouse position calibration table is constructed to meet the expected signal requirements. In the specific implementation, the predetermined query path of the target personnel is read, and the interference area and calibration strategy in the warehouse position calibration table are matched. In addition, the external interference factor is collected through the multivariate monitoring array, and the predetermined interference calibrator is used to analyze it to generate an optimized calibration strategy. The method also includes processing the real-time location information of the RFID reader and writer, and realizing the adaptive parameter calibration of the RFID antenna through the embedded Beidou positioning chip. These technical effects jointly solve the technical problems that the traditional RFID antenna calibration method cannot quickly adapt to environmental changes, resulting in low efficiency and poor accuracy of RFID antenna calibration, and realize the construction of a parameter calibration scheme, and dynamically adjust the calibration strategy based on real-time positioning information, reduce the calibration analysis time, and improve the timeliness and accuracy of the adaptive calibration of the RFID antenna in a complex environment.

[0050] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. The processes depicted in the accompanying drawings do not necessarily require the specific order and continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0051] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0052] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. An adaptive calibration method for RFID tag antenna based on complex environment recognition, characterized in that: include: Collecting attribute features of the target warehouse, simulating and constructing a three-dimensional warehouse space based on the attribute features, simulating RFID signal interference in the three-dimensional warehouse space, and clustering to obtain multiple signal interference areas, wherein obstacle interferences in the same signal interference area are considered the same; For the purpose of meeting expected signal requirements, respectively, performing control parameter calibration analysis of the RFID antenna on the multiple signal interference areas, determining multiple parameter calibration strategies, and constructing a warehouse location calibration table based on the multiple signal interference areas and the multiple parameter calibration strategies; Reading a predetermined query path of a target person, matching the predetermined query path with the location information in the warehouse location calibration table, and determining a plurality of target interference areas and a plurality of matching parameter calibration strategies; The plurality of external interference factors of the plurality of target interference regions are read by a multivariate monitoring array, the plurality of matching parameter calibration strategies and the plurality of external interference factors are input into a predetermined interference calibrator for analysis, a plurality of optimized calibration strategies are output, and an optimized parameter calibration scheme is generated in combination with the plurality of target interference regions; The real-time location information of the RFID reader is obtained, the real-time location information is input into the optimized parameter calibration scheme, and a real-time calibration strategy is output to perform parameter adaptive calibration on the RFID antenna, wherein the RFID reader is embedded with a Beidou positioning chip, and the positioning accuracy meets the expected indicators.

2. The RFID tag antenna adaptive calibration method based on complex environment recognition according to claim 1 is characterized in that: The attribute characteristics of the target warehouse include building structure characteristics, building layout characteristics, building size characteristics, building material characteristics and cargo storage characteristics.

3. The RFID tag antenna adaptive calibration method based on complex environment recognition according to claim 1 is characterized in that: Get multiple signal interference areas, including: Dividing the three-dimensional warehouse space into equal parts according to a predetermined grid size to determine a plurality of warehouse grid spaces; Perform multiple RFID signal interference simulations in the multiple warehouse grid spaces, and obtain multiple signal quality evaluation results after averaging, wherein the signal quality evaluation results at least include signal strength, reading range, and reading rate; Based on the multiple signal quality evaluation results, cluster analysis is performed on the multiple warehouse grid spaces to determine the multiple signal interference areas.

4. The RFID tag antenna adaptive calibration method based on complex environment recognition according to claim 3 is characterized in that: Performing cluster analysis on the multiple warehouse grid spaces, including: Randomly selecting a first grid space and a second grid space from the plurality of warehouse grid spaces, wherein the first grid space and the second grid space are any two adjacent grid spaces from the plurality of warehouse grid spaces; Acquire a first signal strength, a first reading range, and a first reading rate of the first grid space, and a second signal strength, a second reading range, and a second reading rate of the second grid space; If the deviation value between the first signal strength and the second signal strength is less than a predetermined strength threshold, then calculating the deviation between the first reading range and the second reading range; if the range deviation value is less than a predetermined range threshold, then calculating the deviation between the first reading rate and the second reading rate to obtain a reading rate deviation value; If the reading rate deviation value is less than a predetermined reading rate threshold, merging the first grid space and the second grid space to obtain a first merged space, wherein the signal quality evaluation result of the first merged space is the average of the signal quality evaluation results of the first grid space and the second grid space; Iterative merging is continuously performed until there is no adjacent grid space that meets the merging condition, and the multiple signal interference areas are output based on the merging result.

5. The RFID tag antenna adaptive calibration method based on complex environment recognition according to claim 1 is characterized in that: Determine multiple parameter calibration strategies including: Randomly selecting a first signal interference area from the multiple signal interference areas, obtaining a first signal quality evaluation result mean of the first signal interference area, performing a deviation analysis on the first signal quality evaluation result mean based on the expected signal demand, and determining a first signal quality deviation; Taking the RFID device attribute and the first signal quality deviation as constraints and the control parameter calibration of the RFID antenna as the purpose, a plurality of sample calibration data are obtained based on big data retrieval, wherein the sample calibration data includes sample calibration parameters and sample calibration results; Select sample calibration parameters corresponding to sample calibration results that meet expected calibration requirements, construct a standard calibration parameter set, and select the mode of the standard calibration parameter set as the first parameter calibration strategy, which is added to the multiple parameter calibration strategies, wherein the expected calibration requirements include calibration speed, calibration accuracy and calibration stability.

6. The RFID tag antenna adaptive calibration method based on complex environment recognition according to claim 1 is characterized in that: Construct a multi-element monitoring array, including: Configuring external interference indicators, wherein the external interference indicators at least include electromagnetic interference, environmental interference and multipath effect; Based on the external interference index, multiple sensor groups are configured for the multiple target interference areas, and the multi-element monitoring array is constructed by integration.

7. The RFID tag antenna adaptive calibration method based on complex environment recognition according to claim 6 is characterized in that: Build a custom interference calibrator, including: Taking the external interference index as a constraint, retrieving and acquiring a sample calibration data set, performing differential screening on the sample calibration data set, and obtaining a sample training data set, wherein the sample calibration data includes a sample external interference factor, a sample calibration strategy, and a sample optimization calibration strategy; Divide the sample training data set into Q equal parts, select Q parts with replacement, construct a first training set, and iterate and select Q parts Q times to obtain Q training sets; The BP neural network is supervisedly trained and cross-validated using the Q training sets to obtain Q convergence interference calibration units, and the predetermined interference calibrator is constructed based on the Q convergence interference calibration units, wherein the output of the predetermined interference calibrator is the mode of the output results of the Q convergence interference calibration units.

8. The RFID tag antenna adaptive calibration method based on complex environment recognition according to claim 7 is characterized in that: Get a sample training data set, including: Randomly selecting first sample calibration data from the sample calibration data set, wherein the first sample calibration data includes a first sample external interference factor and a first sample calibration strategy; Using a predetermined similarity comparison algorithm, a similarity comparison is performed on the first sample external interference factor and other sample external interference factors in the sample calibration data set, and the number of statistical similarities less than a predetermined factor similarity threshold is set as a first difference degree; If the first difference is greater than a predetermined threshold, a similarity comparison is performed on the first sample calibration strategy and other sample calibration strategies in the sample calibration data set, and the number of similarities less than a predetermined strategy similarity threshold is counted and set as a second difference; If the second difference is greater than a predetermined threshold, the first sample calibration data is added to the sample training data set.

Images