Position resolving method and system for combined processing of vehicle and gun positioning information

Abstract

The invention provides a position calculation method and system for combined processing of vehicle and gun positioning information, and relates to the field of fusion positioning, and the method comprises the steps: determining a plurality of time windows corresponding to vehicle positioning data and gun positioning data based on the size of an optimal time window and an optimal sliding interval; for each time window, determining an alignment reference time point of the time window, generating a vehicle position and a gun position corresponding to the alignment reference time point of the time window, and further calculating a vehicle-gun separation distance corresponding to the alignment reference time point of the time window; determining the vehicle-gun state based on the vehicle-gun separation distance and the dynamic separation threshold value corresponding to the alignment reference time points of the plurality of time windows; when the vehicle-gun state is the fusion state, the vehicle fusion position and the gun fusion position are determined based on the gun positions and the vehicle positions corresponding to the alignment reference time points of the multiple time windows, and the method has the advantage that the accuracy of combined positioning of the vehicle and the gun is improved.

Description

Technical Field

[0001] This invention relates to the field of fusion positioning, and in particular to a method and system for calculating the location of vehicles and firearms by jointly processing their positioning information. Background Technology

[0002] In the armed escort industry, companies monitor the real-time location of vehicles and explosive ordnance-proof firearms to ensure the safety of transport operations. They also need to provide early warnings for situations where vehicles and firearms are separated. Currently, escort companies use two independent systems to meet the needs of real-time location monitoring of vehicles and firearms, as well as other business operations.

[0003] The frequency of vehicle location information reporting generally follows the agreed-upon protocol. An onboard smart terminal is installed in the vehicle, transmitting location data via cellular mobile communication network. The platform receives, parses, and displays the location information. Because the vehicle has a fixed power supply, the data reporting frequency is once every 5 seconds.

[0004] Because there is no unified standard for firearm location, each company typically uses its own proprietary standard, transmitting location data via cellular mobile communication networks. The platform receives, parses, and displays this location information. Due to battery life considerations, firearm data is generally reported every 20 seconds. For regulatory purposes, companies usually display vehicle and firearm location information uniformly on electronic map pages. The varying reporting frequencies from different devices cause vehicles and firearms that are actually in the same location to appear in different positions on the electronic map. When vehicles are moving at high speeds, the lower data reporting frequency leads to increased data latency, further exacerbating location errors. For example, when a vehicle is traveling at high speed, the positional deviation of "vehicle and firearm" on the map can reach 300 meters, which is significantly inconsistent with the actual binding status. This also results in a large number of false alarms for vehicle-firearm separation warnings.

[0005] Therefore, there is a need to provide a location calculation method and system that jointly processes vehicle and firearm positioning information to improve the accuracy of joint vehicle and firearm positioning. Summary of the Invention

[0006] This invention provides a method for jointly processing vehicle and firearm positioning information to calculate location, comprising: determining an optimal time window size and an optimal sliding interval; acquiring vehicle positioning data and firearm positioning data; determining multiple time windows corresponding to the vehicle positioning data and firearm positioning data based on the optimal time window size and the optimal sliding interval; for each time window, determining an alignment reference time point; generating the vehicle position corresponding to the alignment reference time point based on the vehicle positioning data; generating the firearm position corresponding to the alignment reference time point based on the firearm position and vehicle position; calculating the vehicle-firearm separation distance corresponding to the alignment reference time point based on the firearm position and vehicle position based on the alignment reference time point; for each time window, determining a dynamic separation threshold; determining the vehicle-firearm state based on the vehicle-firearm separation distance and dynamic separation threshold corresponding to the alignment reference time points of multiple time windows, wherein the vehicle-firearm state is either a fused state or a separated state; when the vehicle-firearm state is a fused state, determining the vehicle fused position and the firearm fused position based on the firearm position and vehicle position corresponding to the alignment reference time points of multiple time windows.

[0007] Further, determining the optimal time window size and optimal sliding interval includes: acquiring multiple sets of sample data, wherein the sample data includes sample vehicle positioning data and sample gun positioning data in a fused state; determining multiple parameter combinations, wherein the parameter combinations include the time window size and sliding interval; for each parameter combination, calculating the fitness value of the parameter combination based on the multiple sets of sample data; and determining the optimal time window size and optimal sliding interval using a genetic algorithm based on the fitness value of each parameter combination.

[0008] Furthermore, based on multiple sets of sample data, the fitness value of the parameter combination is calculated, including: for each set of sample data, based on the parameter combination, determining multiple time windows corresponding to the sample data and the alignment reference time point for each time window; based on the sample data, calculating the gun position and vehicle position corresponding to the alignment reference time point for each time window; calculating the position correlation coefficient between the gun and the vehicle based on the position correlation coefficient between the gun and the vehicle corresponding to the alignment reference time point for each time window; and calculating the fitness value of the parameter combination based on the position correlation coefficient between the gun and the vehicle corresponding to each set of sample data.

[0009] Furthermore, based on vehicle positioning data, the vehicle position corresponding to the alignment reference time point of the time window is generated, including: determining whether there is an actual vehicle position corresponding to the alignment reference time point of the time window in the vehicle positioning data; if so, the actual vehicle position is used as the vehicle position corresponding to the alignment reference time point of the time window; if not, the vehicle position corresponding to the alignment reference time point of the time window is generated based on the actual vehicle positions at multiple time points within the time window using a motion state feature decomposition algorithm.

[0010] Furthermore, based on the actual vehicle positions at multiple time points within the time window, the motion state feature decomposition algorithm is used to generate the vehicle position corresponding to the alignment reference time point of the time window. This includes: calculating the linear trend component and the dynamic correction component based on the actual vehicle positions at multiple time points within the time window; and generating the vehicle position corresponding to the alignment reference time point of the time window based on the linear trend component and the dynamic correction component.

[0011] Further, determining the dynamic separation threshold corresponding to the time window includes: acquiring vehicle speed data corresponding to the time window; and determining the dynamic separation threshold corresponding to the time window based on the vehicle speed data corresponding to the time window.

[0012] Furthermore, based on the vehicle speed data corresponding to the time window, the dynamic separation threshold corresponding to the time window is determined, including: determining multiple speed sub-windows based on the vehicle speed data corresponding to the time window, wherein the times of any two adjacent speed sub-windows do not overlap; for each speed sub-window, calculating the speed fluctuation value corresponding to the speed sub-window and determining the weight corresponding to the sub-window; performing linear regression calculation using the weighted least squares method based on the vehicle speed data corresponding to the time window and the weight corresponding to each sub-window to obtain the speed linear equation corresponding to the time window; and substituting the alignment reference time point of the time window into the speed linear equation corresponding to the time window to obtain the dynamic separation threshold corresponding to the time window.

[0013] Furthermore, based on the vehicle-gun separation distance and dynamic separation threshold corresponding to the alignment reference time points of multiple time windows, the vehicle-gun state is determined, including: for each time window, based on the vehicle-gun separation distance and dynamic separation threshold corresponding to the alignment reference time point of the time window, the vehicle-gun state corresponding to the time window is determined; based on the vehicle-gun state corresponding to each time window, the vehicle-gun separation probability is calculated; based on the vehicle-gun separation probability, the vehicle-gun state is determined.

[0014] Furthermore, based on the gun and vehicle positions corresponding to the alignment reference time points of multiple time windows, the vehicle fusion position and gun fusion position are determined, including: using a weighted average formula, the vehicle fusion position and gun fusion position are determined based on the gun and vehicle positions corresponding to the alignment reference time points of multiple time windows.

[0015] This invention provides a location calculation system that jointly processes vehicle and firearm positioning information. Applying the aforementioned location calculation method for jointly processing vehicle and firearm positioning information, the system includes: a parameter determination module for determining the optimal time window size and optimal sliding interval; a data acquisition module for acquiring vehicle positioning data and firearm positioning data; and a data analysis module for determining multiple time windows corresponding to the vehicle positioning data and firearm positioning data based on the optimal time window size and optimal sliding interval; for each time window, determining the alignment reference time point of the time window; generating the vehicle position corresponding to the alignment reference time point of the time window based on the vehicle positioning data; and generating the time window based on the firearm positioning data. The data analysis module is used to determine the vehicle-gun separation distance corresponding to the alignment reference time point of the time window, based on the gun position and vehicle position of the alignment reference time point of the time window. The data analysis module is also used to determine the dynamic separation threshold based on vehicle positioning data. Furthermore, the data analysis module is used to determine the vehicle-gun state based on the vehicle-gun separation distance and dynamic separation threshold of the alignment reference time points of multiple time windows, where the vehicle-gun state is either a fused state or a separated state. When the vehicle-gun state is a fused state, the data analysis module is also used to determine the vehicle fusion position and the gun fusion position based on the gun position and vehicle position of the alignment reference time points of multiple time windows.

[0016] Compared with existing technologies, the location calculation method and system for joint processing of vehicle and firearm positioning information provided by this invention has at least the following beneficial effects: 1. Based on a defined time window, an alignment reference time point is determined for each window, and corresponding vehicle and weapon positions are generated respectively. This processing method fully considers the time synchronization of data, ensuring the consistency of vehicle and weapon position information in the time dimension. In practical applications, vehicle and weapon positioning data may come from different sensors, resulting in time delay differences. By aligning the reference time point, this time error can be eliminated, making the vehicle and weapon position information more accurate.

[0017] 2. The dynamic separation threshold is not fixed but dynamically adjusted based on the actual data within the time window. The optimal separation distance between the vehicle and the gun varies under different motion states. This method, by analyzing vehicle and gun positioning data, can adapt to various complex scenarios and accurately determine whether the vehicle and gun are in a fused or separated state.

[0018] 3. The vehicle and gun status is determined based on the separation distance and dynamic separation threshold across multiple time windows. Furthermore, the vehicle-gun separation probability is calculated to ultimately determine the status. This multi-layered and comprehensive approach significantly reduces the possibility of misjudgments. Data from a single time window may be affected by random factors, leading to incorrect vehicle-gun status assessments. By comprehensively considering information from multiple time windows and calculating the separation probability, these random errors can be filtered out, making the judgment more stable and reliable. In practical applications, this can effectively reduce the false alarm rate of vehicle-gun separation warnings, avoiding unnecessary actions and resource waste caused by false alarms.

[0019] 4. When the vehicle and gun are in a fused state, this method determines the fused vehicle and gun positions based on the gun and vehicle positions corresponding to alignment reference time points across multiple time windows using a weighted average formula. This weighted averaging method fully considers the reliability and importance of data from different time windows, making the fused position closer to the actual situation. In cases of rapid vehicle movement or fluctuations in positioning data, it effectively smooths the data, reduces the impact of positioning errors on the fused position, and effectively lowers the false alarm rate of vehicle-gun separation warnings. Attached Figure Description

[0020] This specification will be further described by way of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting; in these embodiments, the same reference numerals denote the same structures, wherein: Figure 1 This is a flowchart illustrating a method for jointly processing vehicle and firearm positioning information to calculate a location, according to some embodiments of this specification. Figure 2 This is a flowchart illustrating the process of generating the vehicle position corresponding to the alignment reference time point of the time window according to some embodiments of this specification; Figure 3 This is a schematic diagram of a location calculation system that jointly processes vehicle and firearm positioning information, according to some embodiments of this specification. Detailed Implementation

[0021] To more clearly illustrate the technical solutions of the embodiments in this specification, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are merely some examples or embodiments of this specification. For those skilled in the art, these drawings can be applied to other similar scenarios without creative effort. Unless obvious from the context or otherwise specified, the same reference numerals in the drawings represent the same structures or operations.

[0022] Figure 1This is a flowchart illustrating a method for jointly processing vehicle and firearm positioning information to calculate a location, based on some embodiments of this specification. Figure 1 As shown, a location calculation method that combines vehicle and firearm positioning information may include the following steps.

[0023] Step 110: Determine the optimal time window size and the optimal sliding interval.

[0024] Specifically, it includes: Acquire multiple sets of sample data, including sample vehicle positioning data and sample gun positioning data in a fused state. The sample vehicle positioning data may include the vehicle position and gun position at multiple consecutive time points. Multiple parameter combinations are determined, including the time window size and the sliding interval. The sliding interval represents the length of time the time window moves forward each time. The window slides gradually at this interval, thereby achieving dynamic processing of continuous data. For example, the time window size is 20 seconds, 30 seconds, 40 seconds, 60 seconds, etc., and the sliding interval is 5 seconds, 10 seconds, 15 seconds, etc. For each parameter combination, the fitness value of the parameter combination is calculated based on multiple sets of sample data; The optimal time window size and optimal sliding interval are determined by using a genetic algorithm based on the fitness value of each parameter combination.

[0025] Specifically, in existing technologies, the difference in reporting frequency between vehicles and firearms leads to temporal asynchrony. For example, vehicles report data every 5 seconds, while firearms report data every 20 seconds. This results in dense data points for vehicles and sparse data points for firearms on the timeline. For instance, within a certain period, a vehicle may have multiple data points, while a firearm may only have a few, making it difficult to directly match the data of vehicles and firearms at the same moment. To solve this technical problem, it is necessary to determine the optimal time window size and the optimal sliding interval. By using a sliding time window algorithm, the data of vehicles and firearms are aligned to the alignment reference time point within each time window, thereby establishing a temporal correspondence between the two data.

[0026] In some embodiments, the fitness value of a parameter combination is calculated based on multiple sets of sample data, including: For each set of sample data, based on the parameter combination, multiple time windows corresponding to the sample data and the alignment reference time point of each time window are determined. Based on the sample data, the gun position and vehicle position corresponding to the alignment reference time point of each time window are calculated. Based on the gun position and vehicle position corresponding to the alignment reference time point of each time window, the position correlation coefficient between the gun and the vehicle is calculated. The fitness value of the parameter combination is calculated based on the correlation coefficient between the position of the gun and the vehicle corresponding to each set of sample data.

[0027] Specifically, the alignment reference time point of the time window can be determined based on the following formula: in, The alignment reference time point for the time window, This is the starting time point of the time window. This is the maximum reporting cycle of the equipment, which is the larger of the vehicle location reporting cycle and the gun location reporting cycle.

[0028] You can refer to step 140 to determine the gun and vehicle positions corresponding to the alignment reference time point of the time window, which will not be repeated hereafter.

[0029] By taking the gun position and vehicle position corresponding to the alignment reference time point of each time window as two variables, and substituting them into the correlation coefficient (e.g., Pearson correlation coefficient) calculation formula, the correlation coefficient between the gun and vehicle positions corresponding to the sample data is obtained.

[0030] For each set of sample data, calculate the variance of the gun position and the variance of the vehicle position at multiple time points included in the sample data. Sum the variances of the gun position and the vehicle position to obtain the total variance corresponding to the sample data. The ratio of the total variance corresponding to the sample data to the sum of the total variances corresponding to all sample data is used as the weight of the sample data.

[0031] Based on the weight of each set of sample data, the fitness value of the parameter combination is obtained by weighting the correlation coefficients between the positions of guns and vehicles corresponding to each set of sample data using a weighted average formula.

[0032] First, each parameter combination is considered an "individual," and numerous parameter combinations constitute the initial "population." These individuals each carry different time window sizes and sliding intervals, representing different possibilities in the solution space. Next, based on the fitness values ​​calculated from multiple sets of sample data, each individual is assigned a fitness score. The fitness value reflects the effectiveness of the parameter combination in practical applications; a higher score indicates that the parameter combination better meets our requirements for dynamically processing continuous data. Subsequently, the genetic algorithm enters its core iterative process. Through the "selection" operation, individuals are filtered according to their fitness scores, with individuals with higher fitness having a greater chance of being retained and entering the next generation of the population, simulating the "survival of the fittest" process in nature. The "crossover" operation pairs selected individuals together, exchanging some of their parameter information to generate new individuals, increasing the diversity of the population. The "mutation" operation randomly changes the parameter values ​​of some individuals with a certain probability, further exploring under-explored regions in the solution space. After multiple rounds of selection, crossover, and mutation operations, the population continuously evolves, and the fitness value gradually increases. Finally, the algorithm stops when the preset number of iterations or the fitness value converges. At this point, the parameter combination corresponding to the individual with the highest fitness in the population is the determined optimal time window size and optimal sliding interval, which can achieve dynamic processing of continuous data in the best way.

[0033] Understandably, the position correlation coefficient directly reflects the degree of correlation between the positions of firearms and vehicles within a time window. By calculating this coefficient and using it to calculate the fitness value, the fitness value can comprehensively reflect the tightness of the correlation between vehicle and firearm data after time alignment under the parameter combination. The stronger the correlation, the more accurately the time window and sliding interval settings can capture the temporal correspondence between the two data, which helps to select parameter combinations that can better solve the problem of temporal asynchrony. Considering the weight of the sample data, the weights are determined based on the variance of the positions of firearms and vehicles in the samples and a weighted average is performed. Variance reflects the dispersion of the data, and sample data with larger weights have a greater impact on the overall fitness value. This method of calculating the fitness value makes the results more focused on those samples with obvious and more representative data characteristics, avoiding excessive interference from individual outliers, and ensuring that the determined parameter combination has wider applicability and stability in practical applications. The accurately calculated fitness value provides a reliable evaluation standard for the genetic algorithm. Genetic algorithms perform selection, crossover, and mutation operations based on fitness values. Fitness values ​​can accurately reflect the quality of parameter combinations, enabling the algorithm to search the solution space more efficiently and converge to the optimal solution quickly. This allows the algorithm to determine the optimal time window size and sliding interval, effectively solving the problem of asynchronous time states between vehicle and firearm data.

[0034] Step 120: Obtain vehicle location data and gun location data.

[0035] The vehicle location data and the gun location data can be vehicle location data at multiple time points within the current time period (e.g., the current 3 minutes or 5 minutes), including vehicle information such as vehicle identification, timestamp (milliseconds), latitude and longitude coordinates (lon, lat), speed, and azimuth angle, as well as gun location data, including gun identification, timestamp (milliseconds), latitude and longitude coordinates (lon, lat), and the bound vehicle identification.

[0036] Based on a pre-defined equipment code list, only valid data conforming to business rules is retained, while redundant information from invalid equipment is removed. Raw firearm and vehicle data is converted into a unified, standardized data format, including core fields such as equipment identifier, timestamp, and latitude / longitude coordinates. Based on historical data distribution and business rules, invalid location points such as those with latitude / longitude coordinates outside the business area or abnormal timestamp jumps are removed to ensure data quality.

[0037] Step 130: Based on the optimal time window size and the optimal sliding interval, determine multiple time windows corresponding to the vehicle positioning data and the gun positioning data.

[0038] Specifically, multiple time windows corresponding to vehicle and firearm location data are determined based on the established optimal time window size and optimal sliding interval. This involves using the optimal time window size as a fixed duration and gradually moving the window forward according to the optimal sliding interval, starting from the data's initial moment. Each movement defines a new time window. This results in multiple partially overlapping time windows on the complete data timeline, which are then used for subsequent alignment and other processing of the vehicle and firearm location data within these windows, thus resolving the temporal asynchrony issue caused by differences in reporting frequencies.

[0039] Step 140: For each time window, determine the alignment reference time point of the time window, generate the vehicle position corresponding to the alignment reference time point of the time window based on vehicle positioning data, generate the gun position corresponding to the alignment reference time point of the time window based on gun positioning data, and calculate the vehicle-gun separation distance corresponding to the alignment reference time point of the time window based on the gun position and vehicle position.

[0040] Specifically, the method for determining the alignment reference time point of the time window can be found in the relevant description of step 110, which will not be repeated here.

[0041] Figure 2 This is a flowchart illustrating the process of generating a time window and determining the vehicle position corresponding to the alignment reference time point, as shown in some embodiments of this specification. Figure 2 As shown, in some embodiments, based on vehicle positioning data, the vehicle position corresponding to the alignment reference time point of the time window is generated, including: Determine whether there is an actual vehicle location corresponding to the alignment reference time point of the time window in the vehicle location data. Specifically, determine whether there is an actual vehicle location corresponding to the alignment reference time point of the time window in the vehicle location data, that is, check whether the vehicle has a true reported location information in the data record at the exact moment of the alignment reference time point. If so, the actual vehicle position will be used as the vehicle position corresponding to the alignment reference time point of the time window. If not, the vehicle position corresponding to the alignment reference time point of the time window is generated based on the actual vehicle position at multiple time points within the time window using the motion state feature decomposition algorithm.

[0042] In some embodiments, a motion state feature decomposition algorithm is used to generate the vehicle position corresponding to the alignment reference time point of the time window based on the actual vehicle positions at multiple time points within the time window, including: Based on the actual vehicle positions at multiple time points within a time window, calculate the linear trend component and the dynamic correction component; Based on the linear trend component and the dynamic correction component, the vehicle position corresponding to the alignment reference time point of the time window is generated.

[0043] Specifically, assuming the vehicle moves at a constant velocity in a straight line within a time window, the linear trend component is calculated based on the velocity and azimuth at multiple time points within the time window: in, For linear trend components, The prediction time interval is defined as the interval between the start of the time window and the alignment reference time point of the time window. The time point preceding the alignment reference time point within the time window. The actual vehicle location, In time The vehicle speed at that time For time The cosine of the azimuth angle of the vehicle's direction of motion (unitless, representing the x-axis component of the direction). For time The sine of the azimuth angle of the vehicle's direction of motion (unitless, representing the y-axis component of the direction). For time The azimuth angle at that time.

[0044] Calculate the rate of change of motion state at multiple time points within the time window: Rate of change of velocity: Rate of change of azimuth: in, For time The vehicle speed at that time For time The vehicle speed at that time For time intervals, For time The azimuth angle at that time.

[0045] Define the historical position deviation matrix: in, For time The linear extrapolation value for time, where n is the total number of time points between the start of the time window and the alignment reference time point of the time window where the actual vehicle position is reported. It is a two-dimensional vector representing the direction and distance of the vehicle's actual position deviating from the extrapolated path.

[0046] Based on the weighted mapping of the rate of change of motion state and historical deviation, the dynamic correction component is obtained: in, The mean of historical deviations related to speed. The azimuth-related historical deviation is calculated based on the historical position deviation matrix. To distinguish the effects of speed and steering, the deviation vector needs to be decomposed into components in two directions. The projection onto the extrapolated velocity direction is used as the velocity direction deviation component to calculate the deviation vector. The projection perpendicular to the extrapolated velocity direction is used as the azimuth direction deviation component. A weighted average of the velocity direction deviation components at all time points before the alignment reference time point within the time window is taken to obtain the mean of the velocity-related historical deviation. Similarly, a weighted average of the azimuth direction deviation components at all time points before the alignment reference time point within the time window is taken to obtain the mean of the azimuth-related historical deviation. , Determined by the time decay factor: , This is the attenuation coefficient (an empirical value of 0.1).

[0047] Based on the linear trend component and the dynamic correction component, the vehicle position corresponding to the alignment reference time point of the time window is generated: in, The vehicle position is the alignment reference time point for the time window.

[0048] Understandably, using a dual mechanism to accurately generate vehicle positions at time windows aligned with reference time points effectively improves the completeness and reliability of positioning data. First, it prioritizes checking if there is actually reported positioning information at the alignment reference time point. If so, the real data is used directly, ensuring zero-error transmission of the original information and avoiding deviations that might be introduced by algorithm intervention. When there is no direct positioning data at that moment, a motion state feature decomposition algorithm is used to calculate linear trend components (reflecting uniform motion) and dynamic correction components (capturing nonlinear changes such as acceleration and steering) based on the actual position data at multiple time points within the time window, ultimately synthesizing the virtual position of the alignment reference point. This design achieves complementary advantages between real data and algorithm estimation: direct data retains its original accuracy, while algorithm estimation, by decomposing motion features (vehicle speed, direction) and weighting corrections, can accurately extrapolate the position at missing moments, especially suitable for complex scenarios involving frequent vehicle acceleration, deceleration, or steering. Furthermore, the dynamic correction component, through a weighted mapping of historical deviations and motion states, makes the estimation results more closely match actual motion patterns, avoiding the cumulative errors of simple linear extrapolation in non-uniform motion. Compared to a single data source or simple interpolation methods, it significantly improves the spatiotemporal continuity of positioning data, providing a more stable basic input for subsequent state determination, and can still maintain high data accuracy, especially in scenarios with sparse data or signal loss.

[0049] The method for generating the gun position corresponding to the alignment reference time point of the time window based on gun positioning data is similar to the method for generating the vehicle position corresponding to the alignment reference time point of the time window based on vehicle positioning data, and will not be repeated here.

[0050] Step 150: For each time window, determine the dynamic separation threshold corresponding to the time window.

[0051] Specifically, it includes: Obtain vehicle speed data corresponding to the time window. The vehicle speed data corresponding to the time window can be the vehicle speed at multiple time points within the time window where the reporting time is located. Based on the vehicle speed data corresponding to the time window, the dynamic separation threshold corresponding to the time window is determined.

[0052] In some embodiments, determining the dynamic separation threshold corresponding to a time window based on vehicle speed data corresponding to that time window includes: Based on the vehicle speed data corresponding to the time window, multiple speed sub-windows are determined, wherein the time of any two adjacent speed sub-windows does not overlap; For each speed sub-window, calculate the speed fluctuation value corresponding to the speed sub-window and determine the weight corresponding to the sub-window. For example, calculate the variance of the speed at each time point included in the speed sub-window as the speed fluctuation value corresponding to the speed sub-window, and use the ratio of the speed fluctuation value corresponding to the speed sub-window to the sum of the speed fluctuation values ​​corresponding to all speed sub-windows as the weight corresponding to the sub-window. By employing weighted least squares, a linear regression operation is performed based on the vehicle speed data corresponding to the time window and the weights of each sub-window to obtain the linear equation for the speed corresponding to the time window. Specifically, the weights of the sub-windows are used as the weights of the speed data at the time points included in the sub-window. The linear regression operation is then performed using weighted least squares, based on the vehicle speed data corresponding to the time window and the weights of the speed data at each time point, to obtain the linear equation for the speed corresponding to the time window. ,in, For predicting speed, a and b are fitting parameters; Substituting the alignment reference time point of the time window into the velocity linear equation corresponding to the time window, the dynamic separation threshold corresponding to the time window is obtained.

[0053] For example, the dynamic separation threshold corresponding to the time window can be calculated using the following formula: in, The dynamic separation threshold corresponding to the time window. To achieve the optimal sliding interval, The positioning error compensation value is initialized to 50 meters. If the current environment is oscillating, the positioning error compensation value will be adjusted accordingly. Dynamic adjustments are made based on the mean of the standard deviations of the velocities at each time point included in the velocity sub-window. Increase proportionally ,For example: .

[0054] Step 160: Determine the vehicle-gun status based on the vehicle-gun separation distance and dynamic separation threshold corresponding to the alignment reference time points of multiple time windows.

[0055] The vehicle and gun are in either a fused state or a separated state.

[0056] In some embodiments, step 160 specifically includes: For each time window, the vehicle-gun state corresponding to the time window is determined based on the vehicle-gun separation distance and dynamic separation threshold corresponding to the alignment reference time point of the time window. For example, if the vehicle-gun separation distance corresponding to the alignment reference time point of the time window is greater than the dynamic separation threshold, the vehicle-gun state corresponding to the time window is a separated state; otherwise, it is a fused state. Based on the vehicle and gun state corresponding to each time window, calculate the vehicle and gun separation probability. Calculate the ratio of the number of time windows in which the vehicle and gun are in a separated state to the total number of time windows, and use this ratio as the vehicle and gun separation probability. The vehicle and gun separation probability is used to determine the vehicle and gun state. For example, if the vehicle and gun separation probability is greater than the vehicle and gun separation probability threshold (e.g., 60%), then the vehicle and gun state is determined to be a separated state.

[0057] Understandably, for each time window, a dynamic separation threshold is determined based on the vehicle speed data within that window. First, non-overlapping speed sub-windows are defined, and the speed fluctuation value and weight of each sub-window are calculated. Then, a weighted least squares linear regression is used to obtain the linear equation for speed, leading to the dynamic separation threshold. This method fully considers the speed change characteristics within different time windows, ensuring the threshold accurately reflects the actual situation in the current window, avoiding potential misjudgments that might occur with fixed thresholds, and improving accuracy. The vehicle-gun separation distance at the aligned baseline time point is compared with the dynamic separation threshold to determine the vehicle-gun state for each time window, and then the vehicle-gun separation probability is calculated. The vehicle-gun state is finally determined by comparing the separation probability with a preset threshold. This multi-level judgment logic integrates local (single time window) and overall (all time windows) information, making the vehicle-gun state determination more scientific and reasonable, and reducing interference from accidental factors. During vehicle operation, the vehicle-gun relationship may change due to various factors. This method, by dynamically adjusting the separation threshold, can promptly capture changes in the vehicle-gun state, adapting to different driving scenarios and operating conditions. Whether driving smoothly or in complex road conditions, it can accurately determine whether the vehicle-gun is in a merged or separated state, thereby reducing the false alarm rate.

[0058] Step 170: When the vehicle and gun are in a fused state, determine the vehicle fused position and the gun fused position based on the gun position and vehicle position corresponding to the alignment reference time points of multiple time windows.

[0059] Specifically, it includes: By using a weighted average formula, the vehicle fusion position and the gun fusion position are determined based on the gun and vehicle positions corresponding to the alignment reference time points of multiple time windows.

[0060] Specifically, using a weighted average formula, the vehicle fusion position is obtained by averaging the gun positions corresponding to the alignment reference time points across multiple time windows based on the gun position weights for each time window. The gun position weights for each time window can be determined based on the standard deviation of the gun positions within that time window; the larger the standard deviation, the greater the gun position weight.

[0061] By using a weighted average formula, the vehicle positions corresponding to the alignment reference time points of multiple time windows are weighted and averaged based on the vehicle position weight of each time window to obtain the gun fusion position. The vehicle position weight of each time window can be determined based on the standard deviation of the vehicle positions within that time window; the larger the standard deviation, the greater the vehicle position weight.

[0062] Understandably, when vehicles are traveling at high speeds, the low data reporting frequency increases positioning errors, leading to significant deviations in the positions of the vehicle and the firearm. This step determines the fused positions of the vehicle and the firearm by weighted averaging the position information from multiple time windows aligned with a reference time point, integrating multiple sets of data. Compared to relying on data from a single time point, this effectively smooths out position deviations caused by data latency and random errors, making the final determined fused position closer to the actual position and improving positioning accuracy. Data latency and positioning errors cause a large number of false alarms in vehicle-firearm separation warnings. This step considers the standard deviation of position data from different time windows to set weights during the determination of the fused position. Time windows with larger standard deviations have greater weight, meaning that more attention is paid to time windows with relatively significant position changes and potentially higher data reliability. Through this scientific weighting method, the true state of the vehicle and firearm can be more accurately judged, avoiding erroneous separation judgments due to individual abnormal data or positioning errors, thereby effectively reducing the false alarm rate of vehicle-firearm separation warnings.

[0063] Figure 3 This is a schematic diagram of a location calculation system that jointly processes vehicle and firearm positioning information, as shown in some embodiments of this specification. Figure 3 As shown, a location calculation system that jointly processes vehicle and firearm positioning information may include a parameter determination module, a data acquisition module, and a data analysis module.

[0064] The parameter determination module is used to determine the optimal time window size and the optimal sliding interval; The data acquisition module is used to acquire vehicle location data and gun location data; The data analysis module is used to determine multiple time windows corresponding to vehicle positioning data and gun positioning data based on the optimal time window size and optimal sliding interval. The data analysis module is also used to determine the alignment reference time point of each time window, generate the vehicle position corresponding to the alignment reference time point of the time window based on vehicle positioning data, generate the gun position corresponding to the alignment reference time point of the time window based on gun positioning data, and calculate the vehicle-gun separation distance corresponding to the alignment reference time point of the time window based on the gun position and vehicle position corresponding to the alignment reference time point of the time window. The data analysis module is also used to determine dynamic separation thresholds based on vehicle positioning data; The data analysis module is also used to determine the vehicle-gun state based on the vehicle-gun separation distance and dynamic separation threshold corresponding to the alignment reference time points of multiple time windows, wherein the vehicle-gun state is either a fused state or a separated state. The data analysis module is also used to determine the vehicle fusion position and the gun fusion position based on the gun position and vehicle position corresponding to the alignment reference time points of multiple time windows when the vehicle and gun are in the fusion state.

[0065] A location calculation system that jointly processes vehicle and firearm positioning information can be used to execute a location calculation method that jointly processes vehicle and firearm positioning information, which will not be elaborated here.

[0066] Finally, it should be understood that the embodiments described in this specification are merely illustrative of the principles of the embodiments described herein. Other variations may also fall within the scope of this specification. Therefore, alternative configurations of the embodiments described herein are intended to be illustrative rather than limiting, and should be considered consistent with the teachings of this specification. Accordingly, the embodiments described herein are not limited to those explicitly introduced and described herein.

Claims

1. A method for calculating the location of vehicles and firearms through joint processing of their positioning information, characterized in that, include: Determine the optimal time window size and the optimal sliding interval; Acquire vehicle location data and gun location data; Based on the optimal time window size and the optimal sliding interval, multiple time windows corresponding to vehicle positioning data and gun positioning data are determined. For each time window, determine the alignment reference time point of the time window, generate the vehicle position corresponding to the alignment reference time point of the time window based on vehicle positioning data, generate the gun position corresponding to the alignment reference time point of the time window based on gun positioning data, and calculate the vehicle-gun separation distance corresponding to the alignment reference time point of the time window based on the gun position and vehicle position. For each time window, determine the corresponding dynamic separation threshold; Based on the vehicle-gun separation distance and dynamic separation threshold corresponding to the alignment reference time points of multiple time windows, the vehicle-gun state is determined, where the vehicle-gun state is either a fused state or a separated state. When the vehicle and gun are in a merged state, the vehicle merged position and the gun merged position are determined based on the gun position and vehicle position corresponding to the alignment reference time points of multiple time windows.

2. The location calculation method for joint processing of vehicle and firearm positioning information according to claim 1, characterized in that, Determining the optimal time window size and optimal sliding interval includes: Acquire multiple sets of sample data, wherein the sample data includes sample vehicle positioning data and sample gun positioning data in a fused state; Determine multiple parameter combinations, wherein the parameter combinations include time window size and sliding interval; For each parameter combination, the fitness value of the parameter combination is calculated based on multiple sets of sample data; The optimal time window size and optimal sliding interval are determined by using a genetic algorithm based on the fitness value of each parameter combination.

3. The location calculation method for joint processing of vehicle and firearm positioning information according to claim 2, characterized in that, Based on multiple sets of sample data, the fitness value of parameter combinations is calculated, including: For each set of sample data, based on the parameter combination, multiple time windows corresponding to the sample data and the alignment reference time point of each time window are determined. Based on the sample data, the gun position and vehicle position corresponding to the alignment reference time point of each time window are calculated. Based on the gun position and vehicle position corresponding to the alignment reference time point of each time window, the position correlation coefficient between the gun and the vehicle is calculated. The fitness value of the parameter combination is calculated based on the correlation coefficient between the position of the gun and the vehicle corresponding to each set of sample data.

4. A location calculation method for joint processing of vehicle and firearm positioning information according to any one of claims 1-3, characterized in that, Based on vehicle positioning data, the vehicle position corresponding to the alignment reference time point of the time window is generated, including: Determine if the vehicle location data contains an alignment reference time point corresponding to the actual vehicle location; If so, the actual vehicle position will be used as the vehicle position corresponding to the alignment reference time point of the time window. If not, the vehicle position corresponding to the alignment reference time point of the time window is generated based on the actual vehicle position at multiple time points within the time window using the motion state feature decomposition algorithm.

5. The location calculation method for joint processing of vehicle and firearm positioning information according to claim 4, characterized in that, Based on the actual vehicle positions at multiple time points within a time window, a motion state feature decomposition algorithm is used to generate the vehicle position corresponding to the alignment reference time point of the time window, including: Based on the actual vehicle positions at multiple time points within a time window, calculate the linear trend component and the dynamic correction component; Based on the linear trend component and the dynamic correction component, the vehicle position corresponding to the alignment reference time point of the time window is generated.

6. A location calculation method for joint processing of vehicle and firearm positioning information according to any one of claims 1-3, characterized in that, Determine the dynamic separation threshold corresponding to the time window, including: Obtain vehicle speed data corresponding to the time window; Based on the vehicle speed data corresponding to the time window, the dynamic separation threshold corresponding to the time window is determined.

7. The location calculation method for joint processing of vehicle and firearm positioning information according to claim 6, characterized in that, Based on the vehicle speed data corresponding to the time window, determine the dynamic separation threshold corresponding to the time window, including: Based on the vehicle speed data corresponding to the time window, multiple speed sub-windows are determined, wherein the time of any two adjacent speed sub-windows does not overlap; For each speed sub-window, calculate the speed fluctuation value corresponding to the speed sub-window and determine the weight of the sub-window; By using the weighted least squares method, linear regression is performed based on the vehicle speed data corresponding to the time window and the weights corresponding to each sub-window to obtain the linear equation of speed corresponding to the time window. Substituting the alignment reference time point of the time window into the velocity linear equation corresponding to the time window, the dynamic separation threshold corresponding to the time window is obtained.

8. A method for calculating the location of a vehicle and a firearms by jointly processing location information according to any one of claims 1-3, characterized in that, Based on the vehicle-gun separation distance and dynamic separation threshold corresponding to the alignment reference time points of multiple time windows, the vehicle-gun state is determined, including: For each time window, the vehicle-gun state corresponding to the time window is determined based on the vehicle-gun separation distance and dynamic separation threshold corresponding to the alignment reference time point of the time window. Calculate the vehicle-gun separation probability based on the vehicle-gun state corresponding to each time window; The vehicle and gun status is determined based on the vehicle-gun separation probability.

9. A location calculation method for joint processing of vehicle and firearm positioning information according to any one of claims 1-3, characterized in that, Based on the alignment reference time points corresponding to multiple time windows, the vehicle fusion position and the firearm fusion position are determined, including: By using a weighted average formula, the vehicle fusion position and the gun fusion position are determined based on the gun and vehicle positions corresponding to the alignment reference time points of multiple time windows.

10. A location calculation system that jointly processes vehicle and firearm positioning information, characterized in that, The location calculation method using the joint processing of vehicle and firearm positioning information as described in claim 1 includes: The parameter determination module is used to determine the optimal time window size and the optimal sliding interval; The data acquisition module is used to acquire vehicle location data and gun location data; The data analysis module is used to determine multiple time windows corresponding to vehicle positioning data and gun positioning data based on the optimal time window size and optimal sliding interval. The data analysis module is also used to: for each time window, determine the alignment reference time point of the time window, generate the vehicle position corresponding to the alignment reference time point of the time window based on vehicle positioning data, generate the gun position corresponding to the alignment reference time point of the time window based on gun positioning data, and calculate the vehicle-gun separation distance corresponding to the alignment reference time point of the time window based on the gun position and vehicle position. The data analysis module is also used to determine the dynamic separation threshold based on vehicle positioning data; The data analysis module is also used to determine the vehicle-gun state based on the vehicle-gun separation distance and dynamic separation threshold corresponding to the alignment reference time points of multiple time windows, wherein the vehicle-gun state is either a fused state or a separated state. The data analysis module is also used to determine the vehicle fusion position and the gun fusion position based on the gun position and vehicle position corresponding to the alignment reference time points of multiple time windows when the vehicle and gun are in a fused state.

Images