Pet indoor and outdoor seamless positioning system and method combining Bluetooth + WiFi

Abstract

The present invention discloses a seamless indoor and outdoor positioning system and method for pets combined with Bluetooth + WiFi, which relates to the field of pet management technology and includes the following steps: a fixed circular area is defined around the pet activity area as a signal monitoring area, and within the signal monitoring area, the signal information around WiFi and Bluetooth is obtained in real time through the smart device worn by the pet. The present invention collects signal information in real time through smart devices, and uses feature engineering and machine learning models to intelligently predict and identify the signal density, ensuring that in a signal-dense or unstable environment, the system can timely adjust and increase the position correction frequency, thereby more accurately calculating the actual position of the pet. By dynamically adjusting the correction frequency and optimizing the positioning strategy, the scheme effectively reduces the positioning error caused by factors such as signal interference, reflection and attenuation, reduces the risk of pets getting lost or entering dangerous areas, and provides a higher level of protection for pet safety.

Description

Technical Field

[0001] The present invention relates to the technical field of pet management, and in particular to a seamless indoor and outdoor positioning system and method for pets combining Bluetooth+WiFi. Background Art

[0002] The Bluetooth + WiFi pet indoor and outdoor seamless positioning system is a positioning solution that combines Bluetooth and WiFi technologies, and is designed to achieve real-time positioning and tracking of pets in indoor and outdoor environments. The system deploys Bluetooth low energy (BLE) beacons and WiFi access points indoors, and uses the strength of these signals (RSSI) and other algorithms (such as triangulation positioning, time difference of arrival (TDOA) etc.) to calculate the pet's position in real time. When the pet is in the indoor area, the Bluetooth beacon provides accurate close-range positioning; while in the outdoor area, the WiFi signal provides a larger coverage range, and through the collaboration of multiple WiFi hotspots, the accurate tracking of the pet's position is ensured. The combination of the two can achieve seamless transition, unaffected by the switching of indoor and outdoor environments, and achieve accurate positioning throughout the process. This system is usually combined with smart devices, such as pet collars, to collect and transmit location data, thereby providing pet owners with real-time and accurate positioning information, making it easy to check the location of the pet at any time and enhance the safety of the pet.

[0003] The prior art has the following deficiencies:

[0004] In the prior art, multiple reflections and diffractions of WiFi and Bluetooth signals in signal-dense areas may lead to errors in position calculation, or even complete failure of positioning. Especially when a pet is missing or in an area that cannot be directly observed, if the positioning system cannot provide accurate position information, the pet may not be found for a long time, thereby increasing the risk of getting lost. If position recognition errors occur in emergency situations, such as when a pet approaches a dangerous area such as a pool or road, it may cause the pet to encounter an accident, such as drowning or being hit by a car.

[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0006] The purpose of the present invention is to provide a seamless indoor and outdoor positioning system and method for pets that combines Bluetooth + WiFi. The system collects signal information in real time through intelligent devices, and uses feature engineering and machine learning models to intelligently predict and identify the signal density, ensuring that in a signal-dense or unstable environment, the system can timely adjust and increase the position correction frequency, thereby more accurately calculating the actual position of the pet. By dynamically adjusting the correction frequency and optimizing the positioning strategy, the solution effectively reduces the positioning error caused by factors such as signal interference, reflection and attenuation, reduces the risk of pets getting lost or entering dangerous areas, and provides higher protection for pet safety, so as to solve the problems in the above-mentioned background technology.

[0007] In order to achieve the above object, the present invention provides the following technical solution: a pet indoor and outdoor seamless positioning method combining Bluetooth + WiFi, comprising the following steps:

[0008] A fixed circular area is defined around the pet's activity area as a signal monitoring area. In the signal monitoring area, the smart device worn by the pet obtains real-time signal information about WiFi and Bluetooth around it, which instantly reflects the actual situation of the signal environment.

[0009] The acquired WiFi and Bluetooth signal information is aggregated into an analysis set, key features that can reflect the signal density in the signal monitoring area are extracted from the established signal analysis set, the extracted data is processed through feature engineering technology, the signal density in the signal monitoring area is quantified through the processed data, the signal data processed by feature engineering is input into a pre-trained machine learning model, and the signal density in the signal monitoring area is intelligently predicted and identified through the machine learning model;

[0010] When a high-density signal is detected in the signal monitoring area, the pet's optimal position range at the next moment is calculated based on the prediction results of the machine learning model and the pet's actual position information at the previous moment, and the position correction frequency is dynamically increased according to the density of the signal to enable more frequent position updates and corrections.

[0011] Preferably, the signal monitoring area is set around the current location of the pet, and the size of the area is determined based on the pet's activity range, actual needs and signal coverage.

[0012] Preferably, key features that can reflect the signal density in the signal monitoring area are extracted from the established signal analysis set, wherein the extracted features include the distribution density of WiFi access points in the signal monitoring area and the degree of spatial aggregation of signal points in the signal monitoring area. The distribution density of WiFi access points in the signal monitoring area and the degree of spatial aggregation of signal points in the signal monitoring area are processed by feature engineering technology to generate access point density reference values ​​and signal concentration reference values, respectively. The signal density of the signal monitoring area is quantified by the access point density reference values ​​and the signal concentration reference values.

[0013] Preferably, the access point density reference value and signal concentration reference value after feature engineering processing are input into a pre-trained machine learning model, and a signal density index is generated by the machine learning model. Based on the signal density index, intelligent prediction and identification of the signal density in the signal monitoring area is achieved.

[0014] Preferably, the signal density index generated when the signal density of the signal monitoring area is intelligently predicted by the pre-trained machine learning model is compared and analyzed with the pre-set signal density index threshold to identify the signal monitoring area with high signal density. The specific steps are as follows:

[0015] If the signal density index is greater than the signal density index threshold, the area where the pet is located is determined as a high-density signal monitoring area;

[0016] If the signal density index is less than or equal to the signal density index threshold, the area where the pet is located is determined to be a normal dense signal monitoring area.

[0017] Preferably, the specific steps of processing the distribution density of WiFi access points in the signal monitoring area by feature engineering technology to generate an access point density reference value are as follows:

[0018] By analyzing the distribution of WiFi access points in the signal monitoring area and combining the spatial structure of the signal monitoring area to quantify the signal density, the reference value of access point density is generated. The generated expression is as follows:

[0019]

[0020] ,in: is the reference value of access point density, n is the total number of WiFi access points in the signal monitoring area, For the i The signal strength of each WiFi access point, No. i The distance between the WiFi access point and the pet's current location, is the signal strength weighted index, which controls the contribution of signal strength. is the distance weighted exponent, which controls the attenuation of the distance. For the i The relative coordinates of the spatial position of each access point and the geometric center of the signal monitoring area reflect the distribution of the access points in the area. R is the radius of the signal monitoring area, indicating the spatial range of the pet's activity area. The spatial concentration adjustment parameter controls the influence of the access point location concentration. is an exponential function that smoothes the spatial distribution of access point density.

[0021] Preferably, the specific steps of processing the concentration degree of signal points in space in the signal monitoring area by feature engineering technology to generate a signal concentration degree reference value are as follows:

[0022] First, define the location of the signal point and its signal strength as ,in , It is a signal point k The spatial coordinates in the signal monitoring area, S is the signal strength;

[0023] Then, the spatial density of the signal points is calculated by introducing a weighted attenuation model, and the following formula is used to represent the influence of the distance between the signal points. The specific expression is:

[0024]

[0025] ,in, Indicates signal point k and j The Euclidean distance between k and j represents the index of the signal point, is the attenuation factor, which reflects the speed at which the signal decays as the distance increases;

[0026] Based on the calculation of weighted density, the signal concentration reference value is defined, and the generation expression of the signal concentration reference value is:

[0027]

[0028] ,in, represents the reference value of signal concentration, N is the total number of signal points in the signal monitoring area, is a nonlinear function of signal strength; and A is the area of ​​the signal monitoring region, used for normalization.

[0029] Preferably, when a high-density signal is identified in the signal monitoring area, the optimal position range of the pet at the next moment is calculated based on the prediction result of the machine learning model and the actual position information of the pet at the previous moment, and the position correction frequency is dynamically increased according to the density of the signal. The specific steps are as follows:

[0030] When the signal monitoring area is identified as a high-density signal area, the actual location information of the pet at the previous moment is combined with the current signal data to calculate the optimal location range of the pet at the next moment. The calculation is performed using the following formula. The specific expression is:

[0031]

[0032] ,in: is the pet's position coordinates at the previous moment, is the signal strength at the current moment, is the signal strength at the previous moment, is the time step, representing the progress in time;

[0033] According to the predicted position at the previous moment and the density of the actual signal, the frequency of position correction is automatically increased to ensure more accurate tracking of the pet's movement. The adjustment of the correction frequency is calculated by the following formula, and the calculation expression is:

[0034]

[0035] ,in: is the basic position correction frequency of the system, It is the corrected frequency that is dynamically adjusted according to the signal density. is a constant coefficient that controls the response rate of the correction frequency as the density increases. is the signal density index of the current signal, is the preset signal density index threshold.

[0036] The pet indoor and outdoor seamless positioning system combining Bluetooth + WiFi includes a signal monitoring area demarcation module, a signal acquisition and real-time monitoring module, a signal information aggregation and feature extraction module, a data processing module, a machine learning model prediction module, a position estimation and optimization module, and a dynamic correction frequency adjustment module;

[0037] The signal monitoring area demarcation module demarcates a fixed circular area around the pet's activity area as the signal monitoring area;

[0038] The signal acquisition and real-time monitoring module obtains the signal information of WiFi and Bluetooth in the signal monitoring area through the smart device worn by the pet, and instantly reflects the actual situation of the signal environment;

[0039] The signal information aggregation and feature extraction module aggregates the acquired WiFi and Bluetooth signal information into an analysis set, and extracts key features that can reflect the signal density in the signal monitoring area from the established signal analysis set;

[0040] The data processing module processes the extracted data through feature engineering technology and quantifies the signal density of the signal monitoring area through the processed data;

[0041] The machine learning model prediction module inputs the signal data processed by feature engineering into the pre-trained machine learning model, and uses the machine learning model to intelligently predict and identify the signal density in the signal monitoring area;

[0042] The position estimation and optimization module, when a high-density signal is identified in the signal monitoring area, calculates the optimal position range of the pet at the next moment based on the prediction results of the machine learning model and the actual position information of the pet at the previous moment;

[0043] The dynamic correction frequency adjustment module dynamically increases the position correction frequency according to the density of the signal, so as to perform position updates and corrections more frequently.

[0044] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0045] The present invention can significantly improve the positioning accuracy and reliability of the signal monitoring area by combining Bluetooth and WiFi to seamlessly locate pets indoors and outdoors. The method collects signal information in real time through intelligent devices, and uses feature engineering and machine learning models to intelligently predict and identify the signal density, ensuring that in a signal-dense or unstable environment, the system can timely adjust and increase the position correction frequency, thereby more accurately calculating the actual position of the pet. By dynamically adjusting the correction frequency and optimizing the positioning strategy, the solution effectively reduces the positioning error caused by factors such as signal interference, reflection and attenuation, reduces the risk of pets getting lost or entering dangerous areas, and provides higher protection for pet safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0047] Figure 1 This is a flow chart of the method for seamless indoor and outdoor positioning of pets combined with Bluetooth + WiFi of the present invention.

[0048] Figure 2This is a module schematic diagram of the pet indoor and outdoor seamless positioning system combining Bluetooth + WiFi of the present invention. DETAILED DESCRIPTION

[0049] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of the present disclosure will be more comprehensive and complete, and the concept of the example embodiments will be fully conveyed to those skilled in the art.

[0050] The present invention provides Figure 1 The pet indoor and outdoor seamless positioning method combining Bluetooth + WiFi shown includes the following steps:

[0051] Designate a fixed circular area around the pet's activity area as a signal monitoring area;

[0052] The signal monitoring area is set around the pet's current location, and the area size is determined based on the pet's activity range, actual needs, and signal coverage;

[0053] The setting of the signal monitoring area is the basis for solving the signal density problem. By limiting the monitoring area, the surrounding signal information related to the pet's location can be concentrated to avoid interference from too many irrelevant areas. The role of this step is to provide a clear area range for subsequent signal collection and processing, ensure that the acquired data is related to the pet's real-time location, and reduce unnecessary external interference.

[0054] The activity range of a pet refers to the size of the area where the pet usually moves in daily life. For example, indoors, a pet may move freely between areas such as the living room, kitchen, and bedroom; while outdoors, the pet's activity range may include open spaces such as courtyards and parks. Setting the size of the signal monitoring area according to the pet's activity habits and activity distance can ensure that the monitoring area covers all areas where the pet may move, thereby effectively tracking the pet's location and signal changes.

[0055] Actual needs refer to adjusting the size of the monitoring area according to the special needs of pets. For example, if the pet is the type that often gets lost or may enter dangerous areas (such as busy streets or environments with pools), a larger monitoring area is needed to obtain signal changes in advance and issue warnings. If the pet's activity space is relatively fixed and it is not easy to leave a specific area, the signal monitoring area can be appropriately reduced. Actual needs may also be related to the positioning accuracy desired by the user. When the accuracy requirement is high, the monitoring area will be correspondingly smaller to enhance the accuracy of positioning.

[0056] Signal coverage refers to the effective propagation range of WiFi and Bluetooth signals. Signal coverage is affected by environmental factors, device power, obstacles, etc. In an environment with weak or unstable signals (such as an indoor environment with many walls), the effective propagation range of the signal is smaller; while in an environment with strong signals (such as an area with densely distributed signal sources), the signal propagation range is larger. Therefore, when setting the size of the monitoring area, it is necessary to adjust the area size according to the actual signal coverage to ensure that the signal in the area can be stably obtained and effectively support the work of the positioning system.

[0057] In the signal monitoring area, the smart device worn by the pet can obtain the signal information of WiFi and Bluetooth in real time, and reflect the actual situation of the signal environment immediately;

[0058] Signal information includes signal strength (RSSI), signal-to-noise ratio (SNR), and other features that may affect positioning accuracy. This process is achieved through smart devices worn by pets (such as pet collars) or fixed access points around them (such as WiFi hotspots, Bluetooth beacons). The acquisition of these signal information is the basis for subsequent analysis and processing, and can reflect the actual situation of the signal environment. The purpose of this step is to collect accurate and real-time signal data as important input data for analyzing signal density and pet location.

[0059] The acquired WiFi and Bluetooth signal information is aggregated into an analysis set, key features that can reflect the signal density in the signal monitoring area are extracted from the established signal analysis set, the extracted data is processed through feature engineering technology, and the signal density of the signal monitoring area is quantified through the processed data;

[0060] Key features that can reflect the signal density in the signal monitoring area are extracted from the established signal analysis set, wherein the extracted features include the distribution density of WiFi access points in the signal monitoring area and the degree of spatial aggregation of signal points in the signal monitoring area. The distribution density of WiFi access points in the signal monitoring area and the degree of spatial aggregation of signal points in the signal monitoring area are processed by feature engineering technology to generate access point density reference values ​​and signal aggregation reference values, respectively. The signal density in the signal monitoring area is quantified by the access point density reference values ​​and the signal aggregation reference values.

[0061] The high density of WiFi access points in the signal monitoring area where the pet is located usually means that the signal density in the signal monitoring area is high. The dense distribution of WiFi access points usually means that the signal coverage in the area is wide and there are more signal sources, especially in large buildings, office areas or commercial areas, where the number and distribution density of WiFi access points are high. In this environment, signals may overlap and multipath effects may occur, resulting in increased signal interference, which makes the positioning system face higher challenges and may affect the positioning accuracy. Therefore, the density of WiFi access points is closely related to the signal density. A high density of WiFi access points usually indicates a high signal density. Conversely, a low signal density may appear in areas with poor signal coverage.

[0062] The specific steps for processing the distribution density of WiFi access points in the signal monitoring area through feature engineering technology to generate access point density reference values ​​are as follows:

[0063] The access point density reference value is generated by analyzing the distribution of WiFi access points in the signal monitoring area and combining it with the spatial structure of the signal monitoring area to quantify the signal density. The generation process is described by the following formula:

[0064]

[0065] ,in: is the reference value of access point density, n is the total number of WiFi access points in the signal monitoring area, For the i The signal strength of a WiFi access point can be obtained through RSSI (Received Signal Strength Indicator). No. i The distance between the WiFi access point and the pet's current location, is a signal strength weighted index that controls the contribution of signal strength (usually greater than 1, indicating that stronger signals contribute more to the density index). is the distance weighted exponent, which controls the attenuation of the distance (usually a positive number, the longer the distance, the smaller the contribution). For the i The relative coordinates of the spatial position of each access point and the geometric center of the signal monitoring area reflect the distribution of the access points in the area. R is the radius of the signal monitoring area, indicating the spatial range of the pet's activity area. The spatial concentration adjustment parameter controls the influence of the access point location concentration. It is an exponential function that smoothes the spatial distribution of access point density to avoid too small density contribution when the access point is too far from the region boundary;

[0066] The location coordinates of the access point relative to the geometric center of the signal monitoring area. Specifically, the geometric center of the signal monitoring area refers to the "center point" of the area, usually the center of all area boundaries. The location of the access point refers to the specific geographical location of each WiFi access point in the area. By calculating the relative coordinates of each access point relative to the geometric center, the offset of the access point relative to the center of the area can be determined, that is, the distance and direction of the access point's position in space relative to the center point. This information helps to understand the spatial characteristics of the access point distribution, and then analyze the signal density and coverage effect in the area.

[0067] Signal strength weighting :This item is measured by signal strength and distance Calculate the contribution of each access point to the signal density. The contribution of signal strength to access point density gradually decreases with increasing distance. However, when the access point is close, the signal strength plays a decisive role, so a weighted index is used to enhance the influence of signal strength. ;

[0068] Spatial distribution regulation : This term smoothes the spatial distribution of access points in the region through an exponential function, taking into account the distance of the access points from the center of the region and the concentration of the distribution. The closer the access point is to the center of the region, the greater the contribution, helping to more accurately reflect the concentration of signals in dense areas;

[0069] Finally, the contributions of all access points are weighted and summed to generate the access point density reference value. The access point density reference value can comprehensively consider multiple factors such as signal strength, distance, spatial distribution, etc., and accurately reflect the signal density in the signal monitoring area;

[0070] The generation logic of the access point density reference value can effectively solve the complex situation where the signal density varies greatly and the access point locations are unevenly distributed within the signal monitoring area, providing a multi-dimensional and dynamic signal density quantification tool.

[0071] It can be seen from the access point density reference value that the larger the performance value of the access point density reference value generated by processing the distribution density of WiFi access points in the signal monitoring area through feature engineering technology, it usually indicates that the signal density in the signal monitoring area is higher. This is because the access point density reference value is calculated by considering factors such as signal strength, access point location and distance. When the distribution density of WiFi access points in the signal monitoring area is high, each access point covers a wider area and has a higher signal strength, resulting in a more concentrated and dense signal. The value of the access point density reference value reflects this, so the larger the value, the higher the coverage and strength of the signal-dense area, and vice versa, it means that the signal coverage in the area is weaker, the access point density is lower, and the signal density is lower. Through the access point density reference value, the dense and sparse areas of the signal can be effectively quantified and distinguished.

[0072] A high degree of aggregation of signal points in space usually indicates that the distribution of signal sources in the signal monitoring area is relatively concentrated, and the signal density is relatively high. This is because the signals of multiple signal sources or access points overlap and enhance in a specific area, resulting in a more concentrated and stable signal strength in that area. Therefore, if the signal point aggregation degree in the signal monitoring area where the pet is located is relatively high, it means that the wireless signal in the area is relatively dense, and may be covered and interfered by more devices, resulting in higher signal quality and positioning accuracy. On the contrary, a low degree of signal point aggregation indicates that the signal distribution in the area is relatively sparse, the signal strength and stability are poor, and the positioning accuracy may be low. Therefore, by monitoring the degree of aggregation of signal points, the signal density of the area where the pet is located can be effectively judged.

[0073] The specific steps of processing the spatial aggregation degree of signal points in the signal monitoring area by feature engineering technology to generate a signal aggregation degree reference value are as follows:

[0074] First, define the location of the signal point and its signal strength as ,in , It is a signal point k The spatial coordinates in the signal monitoring area, S is the signal strength;

[0075] Then, the spatial density of the signal points is calculated by introducing a weighted attenuation model, and the influence of the distance between the signal points is expressed by the following formula:

[0076]

[0077] ,in, Indicates signal point k and j The Euclidean distance between k and jrepresents the index of the signal point, is the attenuation factor (usually ), reflecting the speed at which the signal attenuates as the distance increases. This formula is used to calculate the weighted influence around each signal point, with closer signal points being given higher weights;

[0078] Based on the calculation of weighted density, the signal concentration reference value is defined, and the generation expression of the signal concentration reference value is:

[0079]

[0080] ,in, represents the reference value of signal concentration, N is the total number of signal points in the signal monitoring area, is a nonlinear function of signal strength, for example, or , which is used to enhance the contribution of points with larger signal strength to the aggregation index; A is the area of ​​the signal monitoring area, which is used for normalization;

[0081] The signal concentration reference value finally generated reflects the signal concentration degree of the entire signal monitoring area. The higher the value, the higher the signal density.

[0082] It can be seen from the signal concentration reference value that the larger the performance value of the signal concentration reference value generated by processing the concentration of signal points in space in the signal monitoring area through feature engineering technology, the higher the concentration of signal points in space, that is, multiple signal sources or signal points are concentrated in a certain area, resulting in dense signals in the area, possible signal overlap or interference, thereby enhancing positioning accuracy and stability. On the contrary, a smaller signal concentration reference value indicates that the signal points are sparsely distributed, the signal coverage is uneven, and the signal density is low, which may lead to a decrease in the accuracy of the positioning system. Therefore, the level of the signal concentration reference value directly reflects the density of the signal in the signal monitoring area. The larger the index value, the denser the signal, and vice versa.

[0083] The signal data processed by feature engineering is input into the pre-trained machine learning model, and the machine learning model is used to intelligently predict and identify the signal density in the signal monitoring area;

[0084] The access point density reference value and signal concentration reference value after feature engineering are input into the pre-trained machine learning model, and the signal density index is generated by the machine learning model. Based on the signal density index, intelligent prediction and identification of the signal density in the signal monitoring area are achieved.

[0085] A pre-trained machine learning model is a machine learning model that has been trained with a large amount of historical data and has certain predictive and generalization capabilities. Here, the training process uses labeled data (for example, historical density data of the signal monitoring area) to "teach" the model how to identify and predict signal density from input features. During the training process, the model learns the distribution of signal points, the fluctuation of signal strength, and the changing laws of signal density in different environments, and uses these laws to establish the relationship between input features and signal density.

[0086] During the training process, the machine learning algorithm optimizes a series of parameters so that the model can minimize the error in the training data. Common training algorithms include regression analysis, support vector machine (SVM), decision tree, random forest, etc. These algorithms learn based on different input features (such as access point density reference value and signal concentration reference value), and finally obtain a model that can predict signal density based on new input data. For example, by using historical signal data, the machine learning model can learn that under certain feature values ​​(such as high access point density and signal concentration), the signal density is usually high, and vice versa. Therefore, when facing a new signal monitoring area, the pre-trained model can make predictions based on existing learning experience and automatically determine the location and degree of signal-dense areas.

[0087] Specifically, the pre-trained machine learning model can generate a signal density index based on input features such as access point density reference value and signal concentration reference value. The signal density index represents the signal density of the area. By learning the association between different signal density patterns in the data, the model can intelligently classify or regress the signal monitoring area and output a signal density index. For example, when the input data shows that the distribution of signal points in the area is relatively concentrated and the signal strength is strong, the machine learning model will recognize that the signal density in the area is high and output a higher signal density index. Conversely, when the input data shows that the signal points are relatively scattered and the signal strength is weak, the model will output a lower signal density index.

[0088] The key to this process is to select and process relevant input features (such as access point density and signal concentration) through feature engineering, ensuring that the input data can effectively reflect the characteristics of the signal monitoring area and that the trained machine learning model can handle the complex relationship between these features. With the help of the machine learning model, the system can intelligently predict the signal density based on the input signal information.

[0089] In addition, an important feature of the pre-trained machine learning model is that it can be continuously optimized and adjusted with new data. As more real-time data is input, the model can update its prediction results and further improve the accuracy of signal density prediction. This intelligent prediction mechanism enables the system to dynamically adjust and feedback based on real-time signal data when facing complex and dynamically changing signal environments, ensuring the efficient operation of application scenarios such as pet positioning.

[0090] The machine learning model is not specifically limited here, and can achieve the access point density reference value and signal concentration reference value Perform comprehensive analysis to generate a signal density index In order to realize the technical solution of the present invention, the present invention provides a specific implementation method; the signal density index The generated expression is:

[0091]

[0092] , where , They are the access point density reference values and signal concentration reference value The preset scaling factor of , are all greater than 0. In this formula, the preset proportional coefficient and Indicates the reference value of access point density when calculating the signal density index. and signal concentration reference value Specifically, and are constants set according to actual needs or experience, which determine the influence of the access point density reference value and the signal concentration reference value on the final signal density index. If the value is larger, the reference value of access point density will have a greater impact on the signal density index, and vice versa. By adjusting these preset proportional coefficients, the sensitivity of the model to different features (such as access point density and signal concentration) can be flexibly controlled to ensure that the model can optimize the accuracy of signal density prediction according to different scenarios.

[0093] It can be seen from the signal density index that the larger the performance value of the access point density reference value generated by processing the distribution density of WiFi access points in the signal monitoring area through feature engineering technology, the larger the performance value of the signal concentration reference value generated by processing the degree of concentration of signal points in space in the signal monitoring area through feature engineering technology. That is, the larger the performance value of the signal density index generated when the signal density of the signal monitoring area is intelligently predicted by the pre-trained machine learning model, the higher the signal density in the signal monitoring area, and vice versa.

[0094] The signal density index generated by the pre-trained machine learning model when intelligently predicting the signal density of the signal monitoring area is compared and analyzed with the pre-set signal density index threshold to identify the signal monitoring area with high signal density. The specific steps are as follows:

[0095] If the signal density index is greater than the signal density index threshold, the area where the pet is located is determined as a high-density signal monitoring area;

[0096] If the signal density index is less than or equal to the signal density index threshold, the area where the pet is located is determined to be a normal dense signal monitoring area.

[0097] When a high-density signal is detected in the signal monitoring area, the optimal location range of the pet at the next moment is calculated based on the prediction results of the machine learning model and the actual location information of the pet at the previous moment. The location correction frequency is dynamically increased according to the density of the signal, so that the location can be updated and corrected more frequently.

[0098] When a high-density signal is detected in the signal monitoring area, the optimal location range of the pet at the next moment is calculated based on the prediction results of the machine learning model and the actual location information of the pet at the previous moment, and the location correction frequency is dynamically increased according to the density of the signal. The specific steps are as follows:

[0099] When the signal monitoring area is identified as a high-density signal area, the actual location information of the pet at the previous moment is combined with the current signal data to calculate the optimal location range of the pet at the next moment. The calculation is performed using the following formula. The specific expression is:

[0100]

[0101] ,in: is the pet's position coordinates at the previous moment, is the signal strength at the current moment, is the signal strength at the previous moment, is the time step, which represents the progress in time (usually 1 second or a system-defined time interval);

[0102] The next position of the pet is calculated by combining the current signal density with the position at the previous moment. Affects the predicted range of the pet's position, so that the predicted range of the pet can be dynamically adjusted based on the signal density.

[0103] According to the predicted position at the previous moment and the density of the actual signal, the frequency of position correction is automatically increased to ensure more accurate tracking of the pet's movement. The adjustment of the correction frequency is calculated by the following formula, and the calculation expression is:

[0104]

[0105] ,in: is the basic position correction frequency of the system (for example, once per second), It is the corrected frequency that is dynamically adjusted according to the signal density. is a constant coefficient that controls the response rate of the correction frequency as the density increases. is the signal density index of the current signal, is the preset signal density index threshold;

[0106] when When the dynamic adjustment frequency The frequency of location updates and corrections will increase, ensuring that the system can still accurately track the pet's location in areas with unstable or dense signals.

[0107] By combining the recognition of high-density signal areas with the prediction results of the machine learning model, the location information of the previous moment is used to infer the possible optimal location range of the pet at the next moment, thereby improving the accuracy of the positioning system. In a signal-dense environment, signal fluctuations and errors are more significant, so it is necessary to dynamically adjust the position correction frequency to update the pet's position more frequently to ensure that the positioning error can be corrected in real time when the signal is unstable, thereby avoiding the loss or misjudgment of the pet's position and improving the robustness and response speed of the positioning system.

[0108] By combining Bluetooth and WiFi to seamlessly locate pets indoors and outdoors, the positioning accuracy and reliability of the signal monitoring area can be significantly improved. This method uses smart devices to collect signal information in real time, and uses feature engineering and machine learning models to intelligently predict and identify the signal density, ensuring that in signal-dense or unstable environments, the system can adjust and increase the position correction frequency in a timely manner, thereby more accurately calculating the actual position of the pet. By dynamically adjusting the correction frequency and optimizing the positioning strategy, this solution effectively reduces the positioning error caused by factors such as signal interference, reflection and attenuation, reduces the risk of pets getting lost or entering dangerous areas, and provides higher protection for pet safety.

[0109] The present invention provides Figure 2 The pet indoor and outdoor seamless positioning system combining Bluetooth + WiFi shown includes a signal monitoring area demarcation module, a signal acquisition and real-time monitoring module, a signal information aggregation and feature extraction module, a data processing module, a machine learning model prediction module, a position estimation and optimization module, and a dynamic correction frequency adjustment module;

[0110] The signal monitoring area demarcation module demarcates a fixed circular area around the pet's activity area as the signal monitoring area;

[0111] The signal acquisition and real-time monitoring module obtains the signal information of WiFi and Bluetooth in the signal monitoring area through the smart device worn by the pet, and instantly reflects the actual situation of the signal environment;

[0112] The signal information aggregation and feature extraction module aggregates the acquired WiFi and Bluetooth signal information into an analysis set, and extracts key features that can reflect the signal density in the signal monitoring area from the established signal analysis set;

[0113] The data processing module processes the extracted data through feature engineering technology and quantifies the signal density of the signal monitoring area through the processed data;

[0114] The machine learning model prediction module inputs the signal data processed by feature engineering into the pre-trained machine learning model, and uses the machine learning model to intelligently predict and identify the signal density in the signal monitoring area;

[0115] The position estimation and optimization module, when a high-density signal is identified in the signal monitoring area, calculates the optimal position range of the pet at the next moment based on the prediction results of the machine learning model and the actual position information of the pet at the previous moment;

[0116] The dynamic correction frequency adjustment module dynamically increases the position correction frequency according to the density of the signal, so as to perform position updates and corrections more frequently.

[0117] The method for seamless indoor and outdoor positioning of pets combined with Bluetooth + WiFi provided in an embodiment of the present invention is realized by the above-mentioned seamless indoor and outdoor positioning system for pets combined with Bluetooth + WiFi. The specific methods and processes of the seamless indoor and outdoor positioning system for pets combined with Bluetooth + WiFi are detailed in the embodiment of the method for seamless indoor and outdoor positioning of pets combined with Bluetooth + WiFi, which will not be repeated here.

[0118] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0119] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0120] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0121] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A pet indoor and outdoor seamless positioning method combining Bluetooth + WiFi, characterized in that: The following steps are involved: A fixed circular area is defined around the pet's activity area as a signal monitoring area. In the signal monitoring area, the smart device worn by the pet obtains real-time signal information about WiFi and Bluetooth around it, which instantly reflects the actual situation of the signal environment. The acquired WiFi and Bluetooth signal information is aggregated into an analysis set, key features that can reflect the signal density in the signal monitoring area are extracted from the established signal analysis set, the extracted data is processed through feature engineering technology, the signal density in the signal monitoring area is quantified through the processed data, the signal data processed by feature engineering is input into a pre-trained machine learning model, and the signal density in the signal monitoring area is intelligently predicted and identified through the machine learning model; When a high-density signal is detected in the signal monitoring area, the pet's optimal position range at the next moment is calculated based on the prediction results of the machine learning model and the pet's actual position information at the previous moment, and the position correction frequency is dynamically increased according to the density of the signal to enable more frequent position updates and corrections.

2. The method for seamless indoor and outdoor positioning of pets combining Bluetooth + WiFi according to claim 1, characterized in that: The signal monitoring area is set around the pet's current location, and the area size is determined based on the pet's activity range, actual needs and signal coverage.

3. The method for seamless indoor and outdoor positioning of pets combining Bluetooth + WiFi according to claim 1, characterized in that: Key features that can reflect the signal density in the signal monitoring area are extracted from the established signal analysis set, wherein the extracted features include the distribution density of WiFi access points in the signal monitoring area and the degree of spatial aggregation of signal points in the signal monitoring area. The distribution density of WiFi access points in the signal monitoring area and the degree of spatial aggregation of signal points in the signal monitoring area are processed by feature engineering technology to generate access point density reference values ​​and signal aggregation reference values, respectively. The signal density in the signal monitoring area is quantified by the access point density reference values ​​and the signal aggregation reference values.

4. The method for seamless indoor and outdoor positioning of pets combined with Bluetooth + WiFi according to claim 3, characterized in that: The access point density reference value and signal concentration reference value after feature engineering are input into the pre-trained machine learning model, and the signal density index is generated by the machine learning model. Based on the signal density index, intelligent prediction and identification of the signal density in the signal monitoring area are achieved.

5. The method for seamless indoor and outdoor positioning of pets combined with Bluetooth + WiFi according to claim 4, characterized in that: The signal density index generated by the pre-trained machine learning model when intelligently predicting the signal density of the signal monitoring area is compared and analyzed with the pre-set signal density index threshold to identify the signal monitoring area with high signal density. The specific steps are as follows: If the signal density index is greater than the signal density index threshold, the area where the pet is located is determined as a high-density signal monitoring area; If the signal density index is less than or equal to the signal density index threshold, the area where the pet is located is determined to be a normal dense signal monitoring area.

6. The method for seamless indoor and outdoor positioning of pets combining Bluetooth + WiFi according to claim 3, characterized in that: The specific steps for processing the distribution density of WiFi access points in the signal monitoring area through feature engineering technology to generate access point density reference values ​​are as follows: By analyzing the distribution of WiFi access points in the signal monitoring area and combining the spatial structure of the signal monitoring area to quantify the signal density, the reference value of access point density is generated. The generated expression is as follows: in: is the reference value of access point density, n is the total number of WiFi access points in the signal monitoring area, For the i The signal strength of each WiFi access point, No. i The distance between the WiFi access point and the pet's current location, is the signal strength weighted index, which controls the contribution of signal strength. is the distance weighted exponent, which controls the attenuation of the distance. For the i The relative coordinates of the spatial position of each access point and the geometric center of the signal monitoring area reflect the distribution of the access points in the area. R is the radius of the signal monitoring area, indicating the spatial range of the pet's activity area. The spatial concentration adjustment parameter controls the influence of the access point location concentration. is an exponential function that smoothes the spatial distribution of access point density.

7. The method for seamless indoor and outdoor positioning of pets combined with Bluetooth + WiFi according to claim 3, characterized in that: The specific steps of processing the spatial aggregation degree of signal points in the signal monitoring area by feature engineering technology to generate a signal aggregation degree reference value are as follows: First, define the location of the signal point and its signal strength as ,in , It is a signal point k The spatial coordinates in the signal monitoring area, S is the signal strength; Then, the spatial density of the signal points is calculated by introducing a weighted attenuation model, and the following formula is used to represent the influence of the distance between the signal points. The specific expression is: in, Indicates signal point k and j The Euclidean distance between k and j represents the index of the signal point, is the attenuation factor, which reflects the speed at which the signal decays as the distance increases; Based on the calculation of weighted density, the signal concentration reference value is defined, and the generation expression of the signal concentration reference value is: in, represents the reference value of signal concentration, N is the total number of signal points in the signal monitoring area, is a nonlinear function of signal strength; and A is the area of ​​the signal monitoring region, used for normalization.

8. The method for seamless indoor and outdoor positioning of pets combined with Bluetooth + WiFi according to claim 5, characterized in that: When a high-density signal is detected in the signal monitoring area, the optimal location range of the pet at the next moment is calculated based on the prediction results of the machine learning model and the actual location information of the pet at the previous moment, and the location correction frequency is dynamically increased according to the density of the signal. The specific steps are as follows: When the signal monitoring area is identified as a high-density signal area, the actual location information of the pet at the previous moment is combined with the current signal data to calculate the optimal location range of the pet at the next moment. The calculation is performed using the following formula. The specific expression is: in: is the pet's position coordinates at the previous moment, is the signal strength at the current moment, is the signal strength at the previous moment, is the time step, representing the progress in time; According to the predicted position at the previous moment and the density of the actual signal, the frequency of position correction is automatically increased to ensure more accurate tracking of the pet's movement. The adjustment of the correction frequency is calculated by the following formula, and the calculation expression is: in: is the basic position correction frequency of the system, It is the corrected frequency that is dynamically adjusted according to the signal density. is a constant coefficient that controls the response rate of the correction frequency as the density increases. is the signal density index of the current signal, is the preset signal density index threshold.

9. A pet indoor and outdoor seamless positioning system combining Bluetooth + WiFi, used to implement the pet indoor and outdoor seamless positioning method combining Bluetooth + WiFi as described in any one of claims 1 to 8, characterized in that: It includes a signal monitoring area demarcation module, a signal acquisition and real-time monitoring module, a signal information aggregation and feature extraction module, a data processing module, a machine learning model prediction module, a position estimation and optimization module, and a dynamic correction frequency adjustment module; The signal monitoring area demarcation module demarcates a fixed circular area around the pet's activity area as the signal monitoring area; The signal acquisition and real-time monitoring module obtains the signal information of WiFi and Bluetooth in the signal monitoring area through the smart device worn by the pet, and instantly reflects the actual situation of the signal environment; The signal information aggregation and feature extraction module aggregates the acquired WiFi and Bluetooth signal information into an analysis set, and extracts key features that can reflect the signal density in the signal monitoring area from the established signal analysis set; The data processing module processes the extracted data through feature engineering technology and quantifies the signal density of the signal monitoring area through the processed data; The machine learning model prediction module inputs the signal data processed by feature engineering into the pre-trained machine learning model, and uses the machine learning model to intelligently predict and identify the signal density in the signal monitoring area; The position estimation and optimization module, when a high-density signal is identified in the signal monitoring area, calculates the optimal position range of the pet at the next moment based on the prediction results of the machine learning model and the actual position information of the pet at the previous moment; The dynamic correction frequency adjustment module dynamically increases the position correction frequency according to the density of the signal, so as to perform position updates and corrections more frequently.

Images