Self-adaptive identification method and device for test tube sorting

Abstract

The invention discloses a self-adaptive identification method and device for test tube sorting, and relates to the technical field of sorting, and the method comprises the steps: taking environment and space attributes of test tube sorting based on RTID radio frequency identification as constraints, and determining high-frequency interference factors through radio frequency identification sorting guidance. The scene node set is analyzed by taking minimization of radio frequency interference as an expectation and combining radio frequency identification equipment and test tube attributes to obtain an adaptive parameter threshold. And configuring a monitoring array based on the high-frequency interference factors, obtaining real-time interference factors, inputting the real-time test tube attributes and the interference factors into the scene analysis network for matching, and obtaining real-time adaptive parameters to perform adaptive regulation and control on the radio frequency identification equipment. The technical problems that in an existing test tube sorting technology based on RTID radio frequency identification, complex interference factors are brought by environment and space diversity, so that the radio frequency identification accuracy is low, and the sorting efficiency is low are solved, and the technical effect of improving the radio frequency identification accuracy and the sorting efficiency in test tube sorting is achieved.

Description

Technical Field

[0001] The present application relates to the field of sorting technology, and in particular to an adaptive recognition method and device for test tube sorting. Background Art

[0002] In the test tube sorting scenario based on RTID radio frequency identification, the demand for accurate and efficient test tube sorting is becoming increasingly prominent, and the contradiction in resource demand caused by the influence of interference factors on sorting accuracy is also more prominent. Finding a method that can effectively deal with interference factors to achieve accurate sorting has become a crucial link in solving the test tube sorting problem. Traditional test tube sorting methods are often relatively simple and localized, relying only on fixed radio frequency identification parameters or simple environmental control. They lack overall control over the complex environment and spatial attributes of test tube sorting, lack analysis and response to high-frequency interference factors, and are difficult to fully and accurately evaluate the interference conditions in different scenarios. There are unreasonable situations in parameter adjustment and resource allocation, resulting in low sorting accuracy and low efficiency. The formulation of sorting plans is relatively simple and fixed, and cannot cope well with the changing test tube attributes and interference scenarios.

[0003] Among the current related technologies, there are complex interference factors in the test tube sorting technology based on RTID radio frequency identification due to the diversity of environment and space, and it cannot adapt to the changing environment and test tube properties in real time, resulting in technical problems such as low radio frequency identification accuracy, low sorting efficiency, and poor system stability adaptability. Summary of the invention

[0004] The present application provides an adaptive recognition method and device for test tube sorting, which uses the environment and spatial attributes of test tube sorting based on RTID radio frequency identification as constraints, and uses radio frequency identification sorting to guide the determination of high-frequency interference factors. Based on this, scene enumeration and segmentation are performed, and the adaptation parameter threshold is obtained by combining the radio frequency identification device and the test tube attributes with the expectation of minimizing radio frequency interference. After the mapping is established, a scene parsing tree is built and fused into a network. Based on the high-frequency interference factors, a monitoring array is configured to obtain real-time interference factors, and the real-time test tube attributes and interference factors are input into the scene parsing network for matching, and real-time adaptation parameters are obtained to adaptively control the radio frequency identification device. This achieves the technical effect of improving the accuracy of radio frequency identification and sorting efficiency in test tube sorting by establishing a scene parsing network based on environmental and spatial attributes, and combining real-time monitoring and adaptive regulation.

[0005] The present application provides an adaptive recognition method for test tube sorting, comprising: The environmental attributes and spatial attributes of the test tube sorting are used as associated constraints, and positive sample retrieval is performed under the guidance of radio frequency identification sorting, and high-frequency interference factors are determined according to the sample data set analysis; based on the high-frequency interference factors, scene enumeration is performed to determine multiple interference scenes, and the multiple interference scenes are subdivided according to preset distances to generate multiple scene node sets; with the expectation of minimizing radio frequency interference, the radio frequency identification device and multiple test tube attributes are used as constraints, and the multiple scene node sets are respectively analyzed for radio frequency parameter adaptability to obtain multiple adaptation parameter thresholds, wherein the adaptation parameter thresholds correspond to the test tube attributes one by one, and each adaptation parameter threshold includes multiple adaptation parameter sets; a first mapping of test tube attributes and adaptation parameter thresholds, and a second mapping of scene nodes and adaptation parameters are established, and based on the first mapping and the second mapping, multiple scene parsing trees are built according to the multiple test tube attributes, multiple scene node sets and multiple adaptation parameter thresholds, and a scene parsing network is generated by fusion; a monitoring array is configured based on the high-frequency interference factors, real-time interference factors are obtained at fixed points, real-time test tube attributes and the real-time interference factors are input into the scene parsing network for matching, and real-time adaptation parameters are obtained to perform adaptive regulation on the radio frequency identification device.

[0006] The present application also provides an adaptive recognition device for test tube sorting, comprising: A high-frequency interference factor determination module, the high-frequency interference factor determination module is used to use the environmental attributes and spatial attributes of the test tube sorting as associated constraints, and use radio frequency identification sorting as a guide to perform positive sample retrieval, and determine the high-frequency interference factors based on the sample data set analysis; a scene node set generation module, the scene node set generation module is used to perform scene enumeration based on the high-frequency interference factors to determine multiple interference scenes, and subdivide the multiple interference scenes according to preset distances to generate multiple scene node sets; a radio frequency parameter adaptability analysis module, the radio frequency parameter adaptability analysis module is used to minimize radio frequency interference as the expectation, use radio frequency identification equipment and multiple test tube attributes as constraints, respectively perform radio frequency parameter adaptability analysis on the multiple scene node sets, and obtain multiple adaptation parameter thresholds, Among them, the adaptation parameter threshold corresponds to the test tube attribute one by one, and each adaptation parameter threshold includes multiple adaptation parameter sets; a scene parsing tree building module, the scene parsing tree building module is used to establish a first mapping between the test tube attribute and the adaptation parameter threshold, and a second mapping between the scene node and the adaptation parameter, based on the first mapping and the second mapping, multiple scene parsing trees are built according to the multiple test tube attributes, multiple scene node sets and multiple adaptation parameter thresholds, and a scene parsing network is generated by fusion; an adaptive control module, the adaptive control module is used to configure a monitoring array based on the high-frequency interference factor, obtain real-time interference factors at a fixed point, input the real-time test tube attribute and the real-time interference factor into the scene parsing network for matching, and obtain real-time adaptation parameters to perform adaptive control on the radio frequency identification device.

[0007] The adaptive identification method and device for test tube sorting proposed in this application firstly takes the environment and spatial attributes of test tube sorting based on RTID radio frequency identification as constraints, and guides the determination of high-frequency interference factors with radio frequency identification sorting. Based on this, scene enumeration and segmentation are performed, and the adaptation parameter threshold is obtained by combining the radio frequency identification device and the test tube attributes with the expectation of minimizing radio frequency interference. After the mapping is established, a scene parsing tree is built and fused into a network. Based on the high-frequency interference factors, a monitoring array is configured to obtain real-time interference factors, and the real-time test tube attributes and interference factors are input into the scene parsing network for matching, and real-time adaptation parameters are obtained to adaptively control the radio frequency identification device. By establishing a scene parsing network based on environmental and spatial attributes, and combining real-time monitoring and adaptive control, the technical effect of improving the accuracy of radio frequency identification and sorting efficiency in test tube sorting is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the accompanying drawings of the embodiment of the present invention will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the device according to the embodiment of the present application. It should be understood that the previous or following operations are not necessarily performed precisely in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0009] Figure 1 A schematic diagram of a flow chart of an adaptive recognition method for test tube sorting provided in an embodiment of the present application; Figure 2 A schematic diagram of the structure of an adaptive recognition device for test tube sorting provided in an embodiment of the present application.

[0010] Explanation of the accompanying drawings: high-frequency interference factor determination module 10, scene node set generation module 20, radio frequency parameter adaptability analysis module 30, scene analysis tree construction module 40, adaptive control module 50. DETAILED DESCRIPTION

[0011] 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.

[0012] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0013] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, device, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0014] The present application embodiment provides an adaptive recognition method for test tube sorting, such as Figure 1 As shown, the method includes: Step S100, taking the environmental attributes and spatial attributes of the test tube sorting as the associated constraints, and using RFID sorting as the guide to perform positive sample retrieval, and determine the high-frequency interference factors based on the sample data set analysis. Specifically, determine the environmental attributes of the test tube sorting, such as temperature, humidity, light intensity, etc., and spatial attributes such as the size, layout, and obstacles of the sorting area, and use RFID sorting as the guide to retrieve successful RFID sorting cases from historical data to form a positive sample set, and then conduct an in-depth analysis of the sample data set, classify various factors, observe their performance in the positive samples for different categories of factors, and determine the factors with a high frequency of occurrence and a significant impact on RFID sorting as high-frequency interference factors through statistical analysis.

[0015] In a possible implementation, the environmental attributes and spatial attributes of the test tube sorting are used as associated constraints, and the positive sample retrieval is performed with the radio frequency identification sorting as a guide. The high-frequency interference factors are determined according to the sample data set analysis. Step S100 further includes step S110, which retrieves and obtains the sample data set, wherein the sample data set includes multiple interference factor sets and multiple recognition evaluation sets. Specifically, historical data related to the test tube sorting process is collected through various channels, and the data is organized into a sample data set, wherein the interference factor set includes various factors that may affect the radio frequency identification sorting, such as electromagnetic interference, temperature changes, humidity changes, metal objects approaching, etc., and the recognition evaluation set is an evaluation index of the radio frequency identification effect in each sorting process, such as recognition accuracy, reading time, bit error rate, signal strength, etc.

[0016] Step S120, clustering the multiple interference factor sets to determine multiple sample interference factors and multiple interference frequencies. Specifically, a data clustering algorithm is used to classify the multiple interference factors collected, and similar interference factors are classified into one category by analyzing the characteristics of the interference factors, so as to determine multiple sample interference factors. For example, electromagnetic interference of different intensities are classified into one category, and temperature changes of different degrees are classified into another category, etc. The number of occurrences or proportions of each sample interference factor in the sample data set are counted to determine its interference frequency. For example, the number of occurrences of electromagnetic interference is counted, and its proportion in all samples is calculated as the interference frequency of electromagnetic interference.

[0017] Step S130, according to the multiple interference factor sets and the multiple recognition evaluation sets, the interference intensity of the multiple sample interference factors is calculated to obtain multiple intensity coefficients. Specifically, the interference factor set is used as an independent variable, the recognition evaluation set is used as a dependent variable, based on a single analysis principle, a relationship model between the interference factor and the recognition evaluation is established by using methods such as linear regression fitting, and for each sample interference factor, the degree of its influence on the recognition evaluation is determined by regression analysis to obtain the best fitting coefficient, for example, the influence coefficient of a certain interference factor on the recognition accuracy, the influence coefficient on the reading time, etc. The obtained multiple best fitting coefficients are normalized to make them comparable within a certain range, and finally multiple intensity coefficients are obtained, which reflect the interference intensity of each sample interference factor on the radio frequency identification sorting.

[0018] Step S140, based on the multiple interference frequencies and multiple intensity coefficients, multiple interference frequencies are calculated, and sample interference factors greater than a predetermined frequency threshold are selected as the high-frequency interference factors. Specifically, the interference frequency of each sample interference factor is calculated in combination with the interference frequency and intensity coefficient, and the interference frequency and intensity coefficient are weighted and summed to obtain the interference frequency. The interference frequency comprehensively considers the frequency of occurrence of the interference factor and its interference intensity to the radio frequency identification, sets a predetermined frequency threshold, and determines the sample interference factor with an interference frequency greater than the threshold as a high-frequency interference factor. The high-frequency interference factor is a factor that needs to be focused on and addressed during the test tube sorting process.

[0019] In one possible implementation, based on the multiple interference factor sets and the multiple identification and evaluation sets, the interference intensity of the multiple sample interference factors is calculated to obtain multiple intensity coefficients. Step S130 further includes step S131, taking the interference factor set as the independent variable and the identification and evaluation set as the dependent variable, based on a single analysis principle, performing linear regression fitting according to the multiple interference factor sets and the multiple identification and evaluation sets, determining multiple best fitting coefficients, and normalizing the multiple best fitting coefficients to obtain the multiple intensity coefficients. Specifically, first of all, it is clear that in the analysis process, the set of various interference factors that may affect the radio frequency identification of test tube sorting is taken as the independent variable. The interference factors include electromagnetic interference intensity, temperature change amplitude, humidity level, etc. Each interference factor is an independent variable, and different values ​​are used to represent its different states. The set of evaluation indicators for the radio frequency identification effect is taken as the dependent variable. The recognition evaluation set usually includes recognition accuracy, reading time, bit error rate, signal strength, etc. The indicators reflect the performance and effect of radio frequency identification, and their values ​​will change with the changes of interference factors. The single analysis principle means that in the analysis process, only one dependent variable and the relationship between multiple independent variables are considered each time. For each recognition evaluation indicator (dependent variable), linear regression analysis is performed with multiple interference factors (independent variables) respectively. Using the given multiple interference factor sets and multiple recognition evaluation indicators, the relationship between the dependent variable and the multiple independent variables is analyzed. The data of the price set is used to determine the best fitting coefficient by linear regression fitting. The purpose of linear regression is to find a straight line that minimizes the sum of the distances from all sample points to the straight line. The best coefficient value is determined by minimizing the objective function. For each dependent variable, the value of the objective function is minimized by continuously adjusting the value of the coefficient. After multiple best fitting coefficients are obtained through linear regression fitting, due to the different numerical ranges of the coefficients, in order to facilitate the comparison and analysis of the influence of various interference factors on the dependent variables, the best fitting coefficients need to be normalized. The normalization process can be achieved by mapping the coefficient value to the [0,1] interval or standardizing it to a distribution with a mean of 0 and a variance of 1. The multiple intensity coefficients obtained can more intuitively reflect the influence of each interference factor on the radio frequency identification effect. The larger the intensity coefficient, the greater the influence of the interference factor on the radio frequency identification.

[0020] Step S132, wherein the minimization objective function used for linear regression fitting is: ; where m is the number of samples, is the dependent variable of the ith sample, is the intercept, is the coefficient of the independent variable, n is the number of independent variables, The jth independent variable representing the i-th sample, j is less than or equal to n. Specifically, represents the dependent variable of the ith sample, Represents the value of the dependent variable predicted based on the current coefficient value, by continuously optimizing the coefficient , so that the difference between the predicted value and the actual value is minimized.

[0021] Step S200, based on the high-frequency interference factors, scene enumeration is performed to determine multiple interference scenes, and the multiple interference scenes are subdivided according to a preset distance to generate multiple scene node sets. Specifically, the high-frequency interference factors determined by analyzing the sample data set are clarified, and then various actual situations are enumerated based on these high-frequency interference factors to determine multiple interference scenes, and a preset distance is determined, whether it is a spatial distance or a numerical range, and the multiple enumerated interference scenes are subdivided according to this preset distance, and the sorting area is divided according to the spatial interval or the interference factor value is divided according to a certain interval, so as to generate multiple scene node sets, each scene node represents a specific subdivided interference scene, and contains the interference factor state and related information in the scene.

[0022] Step S300, with the expectation of minimizing radio frequency interference, with the radio frequency identification device and multiple test tube attributes as constraints, respectively perform radio frequency parameter adaptability analysis on the multiple scene node sets, and obtain multiple adaptation parameter thresholds, wherein the adaptation parameter thresholds correspond to the test tube attributes one by one, and each adaptation parameter threshold includes multiple adaptation parameter sets. Specifically, with the expectation of minimizing radio frequency interference, with the radio frequency identification device characteristics and multiple test tube attributes as constraints, respectively perform radio frequency parameter adaptability analysis on multiple scene node sets, by comprehensively considering the interference factors, device characteristics and test tube attributes in the scene, try different radio frequency parameter combinations, determine the most suitable parameters to reduce radio frequency interference and ensure recognition accuracy and stability, and finally obtain multiple adaptation parameter thresholds, which correspond to the test tube attributes one by one, and each adaptation parameter threshold contains multiple adaptation parameter sets, so that in actual applications, according to the specific test tube attributes and interference scenarios, the appropriate parameter set is selected to adjust the parameters of the radio frequency identification device, so as to achieve adaptive radio frequency identification and sorting.

[0023] In one possible implementation, with the expectation of minimizing radio frequency interference and with radio frequency identification equipment and multiple test tube attributes as constraints, radio frequency parameter adaptability analysis is performed on the multiple scene node sets respectively to obtain multiple adaptation parameter thresholds, wherein the adaptation parameter thresholds correspond to the test tube attributes one by one, and each adaptation parameter threshold includes multiple adaptation parameter sets. Step S300 further includes step S310, randomly selecting a first test tube attribute from the multiple test tube attributes, and constructing an radio frequency identification simulation space based on the first test tube attribute and radio frequency identification equipment simulation. Specifically, one of the test tube attributes is randomly selected as the first test tube attribute. The test tube attributes include the material, size, label type, etc. of the test tube. Because different test tube attributes will have different effects on the reflection, absorption and scattering of radio frequency signals, it is necessary to analyze different test tube attributes, and perform simulation based on the selected first test tube attribute and radio frequency identification equipment to construct an radio frequency identification simulation space. In this simulation space, the propagation, reflection and interference of radio frequency signals under different test tube attributes and environmental conditions can be simulated. Through simulation, the influence of radio frequency parameters on radio frequency identification can be analyzed more accurately, providing a basis for subsequent parameter adaptability analysis.

[0024] Step S320, based on the RFID simulation space, with the expectation of minimizing RF interference, perform RF parameter adaptability analysis on the multiple scene node sets respectively, generate a first adaptation parameter threshold, and add it to the multiple adaptation parameter thresholds, wherein the RF parameters include RF frequency, antenna gain, transmission power and transmission rate. Specifically, by using the constructed RFID simulation space, with the desired goal of minimizing RF interference, RF parameter adaptability analysis is performed on multiple scene node sets. The scene node sets represent different interference scenarios and subdivided areas. Different interference factors and environmental conditions may exist in each scene node set. In the analysis process, RF parameters including RF frequency, antenna gain, transmit power and transmission rate are considered. By adjusting these parameters, the changes in RF interference and the accuracy of RF identification are observed. For example, different RF frequencies are tried to see which frequency has the smallest interference and the highest recognition accuracy; the antenna gain is adjusted to observe the changes in signal strength and coverage. By analyzing multiple scene node sets, a first adaptation parameter threshold is generated. The adaptation parameter threshold is the optimal RF parameter range for the first test tube attribute in different scenarios. For example, for a specific test tube material and interference scenario, a suitable RF frequency range, antenna gain range, transmit power range and transmission rate range are determined, and the generated first adaptation parameter threshold is added to multiple adaptation parameter thresholds. With the analysis of different test tube attributes, multiple adaptation parameter thresholds are gradually accumulated to provide more parameter selections for subsequent scene analysis and adaptive regulation.

[0025] In a possible implementation, based on the radio frequency identification simulation space, with the expectation of minimizing radio frequency interference, the radio frequency parameter adaptability analysis is performed on the multiple scene node sets respectively, and a first adaptation parameter threshold is generated, which is added to the multiple adaptation parameter thresholds, wherein the radio frequency parameters include radio frequency frequency, antenna gain, transmission power and transmission rate, and step S320 further includes step S321, constructing a recognition effect evaluation function based on recognition evaluation indicators, wherein the recognition evaluation indicators include recognition accuracy, reading time, bit error rate and signal strength. Specifically, first determine the recognition evaluation indicators, which are crucial to measuring the effect of radio frequency identification. The recognition accuracy reflects the proportion of test tubes that are correctly identified; the reading time indicates the time required to read the test tube label information; the bit error rate reflects the degree of error in the data transmission process; and the signal strength reflects the strength of the radio frequency signal. The recognition effect evaluation function is constructed based on the recognition evaluation indicators, and different weights are assigned to each indicator by weighted summation. The importance of each indicator is determined according to actual needs, and these indicators are combined to quantitatively evaluate the effect of radio frequency identification.

[0026] Step S322, randomly select a first scene node set from the multiple scene node sets, and randomly select a first scene node, use the first scene node to perform scene rendering on the RFID simulation space, and generate a first simulation space. Specifically, randomly select one from the multiple scene node sets as the first scene node set, the scene node set represents different interference scenes and subdivided areas, each scene node set may have different interference factors and environmental conditions, and randomly select a first scene node from the first scene node set, the scene node specifically describes a specific interference situation and environmental characteristics, and use the selected first scene node to perform scene rendering on the RFID simulation space, by adding the interference factors and environmental characteristics in the scene node to the simulation space, so that the simulation space is closer to the actual situation, and the generated first simulation space can more accurately simulate the RFID situation in a specific scene.

[0027] Step S323, read the RF parameter interval of the RFID device, in the first simulation space, take the RF parameter interval as the optimization space, minimize the RF interference as the expectation, perform RF parameter optimization based on the recognition effect evaluation function, obtain the first adaptation parameter, and add it to the first adaptation parameter threshold. Specifically, read the RF parameter interval of the RFID device, the RF parameters include RF frequency, antenna gain, transmission power and transmission rate, etc. These parameters usually have a certain range of values, in the first simulation space, take the RF parameter interval as the optimization space, find the optimal RF parameter combination within this parameter range, minimize the RF interference as the expectation, perform RF parameter optimization based on the recognition effect evaluation function, and adopt various optimization algorithms, such as genetic algorithm. In the genetic algorithm, a set of initial RF parameter combinations are first randomly generated as the population, and the fitness of each parameter combination, that is, the quality of the RF identification effect, is calculated according to the recognition effect evaluation function. Through operations such as selection, crossover and mutation, the population is continuously evolved so that the fitness is gradually improved. After a certain number of iterations, the optimal RF parameter combination is obtained, and the first adaptation parameter obtained is added to the first adaptation parameter threshold. The adaptation parameter is a parameter combination that can enable the RFID device to achieve better performance under specific test tube properties and scenarios. By continuously performing scene rendering and parameter optimization, the first adaptation parameter threshold can be gradually enriched, providing more options for subsequent adaptive regulation.

[0028] Step S400, establish a first mapping of test tube attributes and adaptation parameter thresholds, and a second mapping of scene nodes and adaptation parameters, based on the first mapping and the second mapping, build multiple scene parsing trees according to the multiple test tube attributes, multiple scene node sets and multiple adaptation parameter thresholds, and fuse to generate a scene parsing network. Specifically, establish a first mapping of test tube attributes and adaptation parameter thresholds and a second mapping of scene nodes and adaptation parameters, based on the two mappings, build multiple scene parsing trees with test tube attributes as the starting point, use different test tube attributes as branch nodes, branch expansion is performed according to the scene node set, and then each scene node is connected to the corresponding adaptation parameter, representing the radio frequency parameter adjustment strategy of a specific test tube attribute under different interference scenarios, and finally fuse multiple scene parsing trees to generate a scene parsing network, so that it can comprehensively consider various test tube attributes and interference scenarios, and provide a comprehensive adaptive control solution for radio frequency identification equipment. When real-time test tube attributes and interference factors are input, the adaptation parameters can be quickly found to achieve adaptive control, thereby improving the accuracy and efficiency of test tube sorting.

[0029] In a possible implementation, a first mapping of test tube attributes and adaptation parameter thresholds, and a second mapping of scene nodes and adaptation parameters are established. Based on the first mapping and the second mapping, multiple scene parsing trees are built according to the multiple test tube attributes, multiple scene node sets and multiple adaptation parameter thresholds, and the scenes are fused to generate a scene parsing network. Step S400 further includes step S410. Based on the first mapping and the second mapping, the test tube attribute is used as the first child node, the multiple scene node sets are used as the first leaf node of the first child node, and the adaptation parameter threshold is used as the second leaf node of the first leaf node. Multiple scene parsing trees are built according to the multiple test tube attributes, multiple scene node sets and multiple adaptation parameter thresholds, and the multiple scene parsing trees are fused to generate the scene parsing network. Specifically, based on the first mapping and the second mapping, with the test tube attribute as the first child node, multiple scene node sets as the first leaf node, and the adaptation parameter threshold as the second leaf node of the first leaf node, multiple scenario parsing trees are built according to the multiple test tube attributes, scene node sets and adaptation parameter thresholds. Each scenario parsing tree represents a radio frequency parameter adjustment strategy for a test tube attribute under different interference scenarios. Multiple scenario parsing trees are integrated to generate a scenario parsing network, which can comprehensively consider various test tube attributes and interference scenarios, and provide a comprehensive adaptive control solution for radio frequency identification equipment. When real-time test tube attributes and interference factors are input, the adaptation parameters can be quickly found to achieve adaptive control, thereby improving the accuracy and efficiency of test tube sorting.

[0030] Step S500, configure the monitoring array based on the high-frequency interference factors, obtain the real-time interference factors at a fixed point, input the real-time test tube attributes and the real-time interference factors into the scene analysis network for matching, and obtain the real-time adaptation parameters to adaptively control the radio frequency identification device. Specifically, according to the determined high-frequency interference factors, configure the monitoring array, the array includes various sensors to cover the test tube sorting area, obtain the real-time interference factors at a fixed point, the data processing system analyzes and processes the data collected by the sensors to determine the current real-time interference factors, determine the real-time test tube attributes during the test tube sorting process, input the real-time test tube attributes and the real-time interference factors into the constructed scene analysis network, the scene analysis network will search and match in the network according to the input, determine the real-time adaptation parameters suitable for the current scene, obtain the real-time adaptation parameters to adaptively control the radio frequency identification device, and realize dynamic adaptive control by adjusting device parameters such as radio frequency frequency, antenna gain, etc., to improve the efficiency and accuracy of test tube sorting.

[0031] In a possible implementation, a monitoring array is configured based on the high-frequency interference factors, real-time interference factors are obtained at fixed points, real-time test tube attributes and the real-time interference factors are input into the scene analysis network for matching, and real-time adaptation parameters are obtained to adaptively control the radio frequency identification device. Step S500 further includes step S510, obtaining a historical interference factor monitoring sequence within a preset window. Specifically, a preset time window is first determined. The window can be a period of time in the past, such as the past few minutes, hours, or days. The historical interference factor data within the preset window is collected through the monitoring system to form a historical interference factor monitoring sequence. The interference factors include electromagnetic interference, temperature changes, humidity changes, and other factors that may affect the test tube sorting radio frequency identification. The historical interference factor monitoring sequence records the changes in the interference factors over the past period of time, providing a data basis for subsequent analysis and prediction.

[0032] Step S520, a comprehensive fluctuation analysis is performed based on the historical interference factor monitoring sequence. If the fluctuation trend does not meet the expected index, the interference factor of the next node is predicted according to the fluctuation trend to obtain the predicted interference factor. Specifically, a comprehensive fluctuation analysis is performed on the acquired historical interference factor monitoring sequence, which is achieved by statistical analysis, time series analysis and other methods. For example, the mean, variance, trend and other indicators of the interference factor are calculated to understand the change law and fluctuation of the interference factor, and the expected index is set. These indicators can be determined according to actual needs and experience. For example, the fluctuation range and change speed of the interference factor are set as the expected index. If the analysis finds that the fluctuation trend of the interference factor does not meet the expected index, it means that the change of the interference factor may exceed the normal range, and further prediction is required. According to the fluctuation trend of the interference factor, an appropriate prediction method is used to predict the interference factor of the next node. For example, a time series prediction model, such as an ARIMA model, a neural network model, etc., can be used to predict future interference factors based on historical data. The predicted interference factors are obtained through prediction, which provides a basis for subsequent predictive regulation.

[0033] Step S530, predictive adjustment is performed on the radio frequency identification device of the next node according to the predicted interference factor matching and the predicted adaptation parameters. Specifically, once the predicted interference factors are obtained, the system matches the adaptation parameters according to these prediction results, which is achieved by querying the established adaptation parameter database or using a specific algorithm. The adaptation parameter database stores the corresponding adaptation parameters under different interference factors. The adaptation parameters are determined by the previous solution process, which can enable the radio frequency identification device to achieve better performance in a specific interference scenario. After finding the corresponding predicted adaptation parameters according to the predicted interference factors, predictive adjustment is performed on the radio frequency identification device of the next node, which means that before the interference factors actually occur, the system can adjust the working parameters of the radio frequency identification device in advance to adapt to the interference situation. Predictive adjustment can improve the adaptability of the RFID system in complex and dynamic environments, ensure the efficiency and accuracy of the test tube sorting process, and by adjusting the equipment parameters in advance, the system can better cope with the changes in interference factors, reduce the impact on test tube sorting, and improve the speed and accuracy of sorting.

[0034] The embodiment of the present application uses the environmental and spatial attributes of test tube sorting based on RTID radio frequency identification as constraints, and uses radio frequency identification sorting to guide the determination of high-frequency interference factors. Based on this, scene enumeration and segmentation are performed, and with the expectation of minimizing radio frequency interference, the scene node set is parsed in combination with radio frequency identification equipment and test tube attributes to obtain adaptation parameter thresholds. After the mapping is established, a scene parsing tree is built and fused into a network. Based on the high-frequency interference factors, a monitoring array is configured to obtain real-time interference factors, and the real-time test tube attributes and interference factors are input into the scene parsing network for matching, and real-time adaptation parameters are obtained to perform adaptive regulation of the radio frequency identification device. By establishing a scene parsing network based on environmental and spatial attributes, and combining real-time monitoring and adaptive regulation, the technical effect of improving the accuracy of radio frequency identification in test tube sorting and the sorting efficiency is achieved.

[0035] In the above, refer to Figure 1 A method for adaptive identification of test tube sorting according to an embodiment of the present invention is described in detail. Figure 2 An adaptive recognition device for test tube sorting according to an embodiment of the present invention is described.

[0036] According to an embodiment of the present invention, an adaptive recognition device for test tube sorting is used to solve the technical problems of low accuracy of radio frequency identification, low sorting efficiency, and poor adaptability of system stability caused by complex interference factors brought about by the diversity of environment and space in the existing test tube sorting technology based on RTID radio frequency identification, which cannot adapt to the changing environment and test tube properties in real time. By establishing a scene analysis network based on environment and space properties, and combining real-time monitoring and adaptive regulation, the technical effect of improving the accuracy of radio frequency identification and sorting efficiency in test tube sorting is achieved. An adaptive recognition device for test tube sorting includes: a high-frequency interference factor determination module 10, a scene node set generation module 20, a radio frequency parameter adaptability analysis module 30, a scene analysis tree construction module 40, and an adaptive regulation module 50.

[0037] The high-frequency interference factor determination module 10 is used to perform positive sample retrieval based on the environmental attributes and spatial attributes of the test tube sorting as associated constraints and the radio frequency identification sorting as a guide, and determine the high-frequency interference factors based on the sample data set analysis; The scene node set generation module 20 is used to perform scene enumeration based on the high-frequency interference factors to determine multiple interference scenes, subdivide the multiple interference scenes according to preset distances, and generate multiple scene node sets; The radio frequency parameter adaptability analysis module 30 is used to perform radio frequency parameter adaptability analysis on the multiple scene node sets respectively with the expectation of minimizing radio frequency interference and with the radio frequency identification device and multiple test tube attributes as constraints, and obtain multiple adaptation parameter thresholds, wherein the adaptation parameter thresholds correspond to the test tube attributes one by one, and each adaptation parameter threshold includes multiple adaptation parameter sets; The scene parsing tree building module 40 is used to establish a first mapping between test tube attributes and adaptation parameter thresholds, and a second mapping between scene nodes and adaptation parameters. Based on the first mapping and the second mapping, multiple scene parsing trees are built according to the multiple test tube attributes, multiple scene node sets and multiple adaptation parameter thresholds, and the scene parsing network is generated by fusion. The adaptive control module 50 is used to configure the monitoring array based on the high-frequency interference factors, obtain the real-time interference factors at a fixed point, input the real-time test tube attributes and the real-time interference factors into the scene analysis network for matching, and obtain the real-time adaptation parameters to perform adaptive control on the radio frequency identification device.

[0038] The specific configuration of the high-frequency interference factor determination module 10 will be described in detail below. As described above, the environmental attributes and spatial attributes of the test tube sorting are used as association constraints, and the radio frequency identification sorting is used as a guide to perform positive sample retrieval, and the high-frequency interference factors are determined according to the sample data set analysis. The high-frequency interference factor determination module 10 further includes: a sample data set retrieval unit, the sample data set retrieval unit is used to retrieve and obtain a sample data set, wherein the sample data set includes multiple interference factor sets and multiple identification evaluation sets; a factor clustering unit, the factor clustering unit is used to perform factor clustering on the multiple interference factor sets, determine multiple sample interference factors and multiple interference frequencies; an interference intensity calculation unit, the interference intensity calculation unit is used to perform interference intensity calculation on the multiple sample interference factors according to the multiple interference factor sets and multiple identification evaluation sets, and obtain multiple intensity coefficients; a high-frequency interference factor setting unit, the high-frequency interference factor setting unit is used to calculate multiple interference frequencies based on the multiple interference frequencies and multiple intensity coefficients, and select sample interference factors greater than a predetermined frequency threshold as the high-frequency interference factors.

[0039] Wherein, according to the multiple interference factor sets and the multiple identification evaluation sets, the interference intensity calculation is performed on the multiple sample interference factors to obtain multiple intensity coefficients, and the interference intensity calculation unit further includes: a linear regression fitting subunit, the linear regression fitting subunit is used to take the interference factor set as an independent variable and the identification evaluation set as a dependent variable, based on a single analysis principle, perform linear regression fitting according to the multiple interference factor sets and the multiple identification evaluation sets, determine multiple best fitting coefficients, and normalize the multiple best fitting coefficients to obtain the multiple intensity coefficients; an objective function subunit, the objective function subunit is used in which the minimization objective function used for linear regression fitting is: wherein the minimization objective function used for linear regression fitting is: ; where m is the number of samples, is the dependent variable of the ith sample, is the intercept, is the coefficient of the independent variable, n is the number of independent variables, The j-th independent variable that represents the i-th sample, where j is less than or equal to n.

[0040] The specific configuration of the RF parameter adaptability analysis module 30 will be described in detail below. As described above, with the expectation of minimizing RF interference, with the RFID device and multiple test tube attributes as constraints, the RF parameter adaptability analysis is performed on the multiple scene node sets respectively, and multiple adaptation parameter thresholds are obtained, wherein the adaptation parameter thresholds correspond to the test tube attributes one by one, and each adaptation parameter threshold includes multiple adaptation parameter sets. The RF parameter adaptability analysis module 30 further includes: a simulation space identification unit, the simulation space identification unit is used to randomly select a first test tube attribute from the multiple test tube attributes, and construct an RFID simulation space based on the first test tube attribute and the RFID device simulation; a RF parameter adaptability analysis unit, the RF parameter adaptability analysis unit is used to perform RF parameter adaptability analysis on the multiple scene node sets respectively based on the RFID simulation space, with the expectation of minimizing RF interference, generate a first adaptation parameter threshold, and add it to the multiple adaptation parameter thresholds, wherein the RF parameters include RF frequency, antenna gain, transmission power and transmission rate.

[0041] Among them, based on the radio frequency identification simulation space, with the expectation of minimizing radio frequency interference, the radio frequency parameter adaptability analysis is performed on the multiple scene node sets respectively, and a first adaptation parameter threshold is generated, which is added to the multiple adaptation parameter thresholds, wherein the radio frequency parameters include radio frequency frequency, antenna gain, transmission power and transmission rate, and the radio frequency parameter adaptability analysis unit further includes: a recognition effect evaluation function construction subunit, the recognition effect evaluation function construction subunit is used to construct a recognition effect evaluation function based on recognition evaluation indicators, wherein the recognition evaluation indicators include recognition accuracy, reading time, bit error rate and signal strength; a scene rendering subunit, the scene rendering subunit is used to randomly select a first scene node set from the multiple scene node sets, and randomly select a first scene node, use the first scene node to perform scene rendering on the radio frequency identification simulation space, and generate a first simulation space; a parameter optimization subunit, the parameter optimization subunit is used to read the radio frequency parameter interval of the radio frequency identification device, in the first simulation space, take the radio frequency parameter interval as the optimization space, with the expectation of minimizing radio frequency interference, perform radio frequency parameter optimization based on the recognition effect evaluation function, obtain a first adaptation parameter, and add it to the first adaptation parameter threshold.

[0042] The specific configuration of the scene parsing tree building module 40 will be described in detail below. As described above, a first mapping of the test tube attribute and the adaptation parameter threshold, and a second mapping of the scene node and the adaptation parameter are established. Based on the first mapping and the second mapping, multiple scene parsing trees are built according to the multiple test tube attributes, multiple scene node sets and multiple adaptation parameter thresholds, and the scene parsing network is generated by fusion. The scene parsing tree building module 40 further includes: a scene parsing network fusion unit, the scene parsing network fusion unit is used to build multiple scene parsing trees according to the multiple test tube attributes, multiple scene node sets and multiple adaptation parameter thresholds based on the first mapping and the second mapping, with the test tube attribute as the first child node, the multiple scene node sets as the first leaf node of the first child node, and the adaptation parameter threshold as the second leaf node of the first leaf node, and the multiple scene parsing trees are fused to generate the scene parsing network.

[0043] The specific configuration of the adaptive control module 50 will be described in detail below. As described above, the monitoring array is configured based on the high-frequency interference factors, the real-time interference factors are obtained at a fixed point, the real-time test tube attributes and the real-time interference factors are input into the scene analysis network for matching, and the real-time adaptation parameters are obtained to adaptively control the radio frequency identification device. The adaptive control module 50 further includes: a historical interference factor monitoring sequence acquisition unit, the historical interference factor monitoring sequence acquisition unit is used to obtain the historical interference factor monitoring sequence within a preset window; a comprehensive fluctuation analysis unit, the comprehensive fluctuation analysis unit is used to perform a comprehensive fluctuation analysis based on the historical interference factor monitoring sequence, if the fluctuation trend does not meet the expected indicators, the interference factor of the next node is predicted according to the fluctuation trend to obtain the predicted interference factor; a predicted adaptation parameter unit, the predicted adaptation parameter unit is used to obtain the predicted adaptation parameter according to the predicted interference factor matching, and predictively control the radio frequency identification device of the next node.

[0044] An adaptive recognition device for test tube sorting provided by an embodiment of the present invention can execute an adaptive recognition method for test tube sorting provided by any embodiment of the present invention, and has functional modules and beneficial effects corresponding to the execution method.

[0045] Although the present application makes various references to certain modules in the device according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0046] The above specific implementation manner does not constitute a limitation to the protection scope of the present application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the protection scope of the present application. In some cases, the actions or steps recorded in the present application can be performed in an order different from that in the embodiment and can still achieve the desired results. In addition, the process depicted in the accompanying drawings does not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. An adaptive recognition method for test tube sorting, characterized in that: include: Taking the environmental attributes and spatial attributes of the test tube sorting as the associated constraints, the positive sample retrieval is guided by the RFID sorting, and the high-frequency interference factors are determined based on the sample data set analysis; Perform scene enumeration based on the high-frequency interference factors to determine multiple interference scenes, subdivide the multiple interference scenes according to preset distances, and generate multiple scene node sets; With the expectation of minimizing radio frequency interference and with radio frequency identification devices and multiple test tube attributes as constraints, radio frequency parameter adaptability analysis is performed on the multiple scene node sets respectively to obtain multiple adaptation parameter thresholds, wherein the adaptation parameter thresholds correspond to the test tube attributes one by one, and each adaptation parameter threshold includes multiple adaptation parameter sets; Establishing a first mapping between test tube attributes and adaptation parameter thresholds, and a second mapping between scene nodes and adaptation parameters, and building multiple scene parsing trees based on the multiple test tube attributes, multiple scene node sets, and multiple adaptation parameter thresholds, and fusing and generating a scene parsing network; Based on the high-frequency interference factors, a monitoring array is configured, real-time interference factors are acquired at fixed points, real-time test tube properties and the real-time interference factors are input into the scene analysis network for matching, and real-time adaptation parameters are acquired to perform adaptive regulation on the radio frequency identification device.

2. The adaptive recognition method for test tube sorting according to claim 1, characterized in that: Identify high-frequency interference factors, including: Retrieving and acquiring a sample data set, wherein the sample data set includes a plurality of interference factor sets and a plurality of recognition evaluation sets; Performing factor clustering on the multiple interference factor sets to determine multiple sample interference factors and multiple interference frequencies; According to the multiple interference factor sets and the multiple identification evaluation sets, interference intensity calculation is performed on the multiple sample interference factors to obtain multiple intensity coefficients; A plurality of interference frequencies are calculated based on the plurality of interference frequencies and the plurality of intensity coefficients, and sample interference factors greater than a predetermined frequency threshold are selected as the high-frequency interference factors.

3. The adaptive recognition method for test tube sorting according to claim 2, characterized in that: Get multiple intensity coefficients, including: Taking the interference factor set as an independent variable and the identification evaluation set as a dependent variable, based on a single analysis principle, performing linear regression fitting according to the multiple interference factor sets and the multiple identification evaluation sets, determining multiple best fitting coefficients, and normalizing the multiple best fitting coefficients to obtain the multiple intensity coefficients; Among them, the minimization objective function used for linear regression fitting is: ； Where m is the number of samples, is the dependent variable of the ith sample, is the intercept, is the coefficient of the independent variable, n is the number of independent variables, The j-th independent variable that represents the i-th sample, where j is less than or equal to n.

4. The adaptive recognition method for test tube sorting according to claim 1, characterized in that: Get multiple adaptation parameter thresholds, including: Randomly selecting a first test tube attribute from the multiple test tube attributes, and constructing a radio frequency identification simulation space based on the first test tube attribute and radio frequency identification device simulation; Based on the RFID simulation space, with the expectation of minimizing RF interference, RF parameter adaptability analysis is performed on the multiple scene node sets respectively, and a first adaptation parameter threshold is generated and added to the multiple adaptation parameter thresholds, wherein the RF parameters include RF frequency, antenna gain, transmission power and transmission rate.

5. The adaptive recognition method for test tube sorting according to claim 4, characterized in that: Generating a first adaptation parameter threshold includes: Constructing a recognition effect evaluation function based on recognition evaluation indicators, wherein the recognition evaluation indicators include recognition accuracy, reading time, bit error rate and signal strength; Randomly selecting a first scene node set from the multiple scene node sets, and randomly selecting a first scene node, and using the first scene node to perform scene rendering on the RFID simulation space to generate a first simulation space; The radio frequency parameter interval of the radio frequency identification device is read, and in the first simulation space, the radio frequency parameter interval is used as the optimization space, and the radio frequency interference is minimized as the expectation, and the radio frequency parameter optimization is performed based on the recognition effect evaluation function to obtain the first adaptation parameter, and add it to the first adaptation parameter threshold.

6. The adaptive recognition method for test tube sorting according to claim 1, characterized in that: Generate a scene parsing network, including: Based on the first mapping and the second mapping, with the test tube attribute as the first child node, multiple scene node sets as the first leaf node of the first child node, and the adaptation parameter threshold as the second leaf node of the first leaf node, multiple scene parsing trees are built according to the multiple test tube attributes, multiple scene node sets and multiple adaptation parameter thresholds, and the multiple scene parsing trees are merged to generate the scene parsing network.

7. The adaptive recognition method for test tube sorting according to claim 1, characterized in that: Obtaining real-time adaptation parameters to adaptively control the RFID device, and then also including: Obtaining the historical interference factor monitoring sequence within a preset window; Performing a comprehensive fluctuation analysis based on the historical interference factor monitoring sequence, if the fluctuation trend does not meet the expected index, predicting the interference factor of the next node according to the fluctuation trend to obtain the predicted interference factor; According to the predicted interference factor matching, the predicted adaptation parameters are obtained, and the radio frequency identification device of the next node is predictively regulated.

8. An adaptive recognition device for test tube sorting, characterized in that: The device is used to implement an adaptive recognition method for test tube sorting according to any one of claims 1 to 7, and the device comprises: A high-frequency interference factor determination module, which is used to perform positive sample retrieval based on the environmental attributes and spatial attributes of the test tube sorting as association constraints and the radio frequency identification sorting as a guide, and determine the high-frequency interference factors based on the sample data set analysis; A scene node set generation module, the scene node set generation module is used to perform scene enumeration based on the high-frequency interference factors to determine multiple interference scenes, subdivide the multiple interference scenes according to preset distances, and generate multiple scene node sets; A radio frequency parameter adaptability analysis module, wherein the radio frequency parameter adaptability analysis module is used to perform radio frequency parameter adaptability analysis on the multiple scene node sets respectively with the expectation of minimizing radio frequency interference and with radio frequency identification devices and multiple test tube attributes as constraints, and obtain multiple adaptation parameter thresholds, wherein the adaptation parameter thresholds correspond to the test tube attributes one by one, and each adaptation parameter threshold includes multiple adaptation parameter sets; A scene parsing tree building module, the scene parsing tree building module is used to establish a first mapping between test tube attributes and adaptation parameter thresholds, and a second mapping between scene nodes and adaptation parameters, based on the first mapping and the second mapping, build multiple scene parsing trees according to the multiple test tube attributes, multiple scene node sets and multiple adaptation parameter thresholds, and fuse them to generate a scene parsing network; The adaptive control module is used to configure the monitoring array based on the high-frequency interference factors, obtain the real-time interference factors at a fixed point, input the real-time test tube attributes and the real-time interference factors into the scene analysis network for matching, and obtain the real-time adaptation parameters to perform adaptive control on the radio frequency identification device.

Images