Multi-person riding detection method and device, vehicle, electronic equipment and storage medium

Abstract

The invention provides a multi-person riding detection method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining a reference weight parameter of a user in response to a vehicle use request of the user for a target vehicle, the reference weight parameter being determined based on historical riding data of the user; multi-person riding detection is performed based on the real-time load parameter of the target vehicle and the reference weight parameter, a multi-person riding detection result indicating whether the target vehicle currently has a multi-person riding behavior is obtained, and the personalized reference weight parameter based on historical riding data is introduced, so that the limitation of a fixed threshold is broken, and the riding experience of the target vehicle is improved. Accurate discrimination can be carried out based on historical weight characteristics of each user, the problem that a large-weight user is easily misjudged as double-person riding so that the user cannot use the bicycle is effectively solved, the problem of missed discrimination caused by the fact that the total weight does not exceed a fixed threshold value when two small-weight users ride the bicycle at the same time is effectively solved, and on the premise that hardware cost is not greatly increased, the user experience is improved. And the detection accuracy and robustness are greatly improved.

Description

Technical Field

[0001] This invention relates to the field of vehicle management technology, and in particular to a method, device, vehicle, electronic device, and storage medium for detecting multiple riders. Background Technology

[0002] With the booming development of the sharing economy, shared electric bicycles have become widely popular as a green and convenient short-distance travel tool. However, the phenomenon of multiple riders persists despite repeated bans. This behavior seriously affects the vehicle's handling balance, increases braking distance, raises the risk of accidents, violates traffic regulations, and poses hidden dangers to user safety and business operations.

[0003] Currently, the primary method for determining whether multiple riders are present is by installing gravity sensors on the vehicle to detect changes in load, such as the amount of load change within a preset time period or whether the current load exceeds a preset maximum load. However, this method has significant risks of missed detections and false alarms in practical applications. Specifically, if two lighter users board simultaneously, the sensor may struggle to detect the secondary weight gain. If a maximum load is set to avoid false alarms, such two-person riding will not be identified. Conversely, if the maximum load is lowered to avoid missed detections, a legally registered heavier user may be falsely flagged as overloaded. In short, the current solution fails to meet the needs of users with different weights, resulting in a significant contradiction. Summary of the Invention

[0004] This invention provides a method, device, vehicle, electronic device, and storage medium for detecting multiple riders, in order to solve the problem that the existing technology cannot meet the needs of users with different weights, and to improve the accuracy and robustness of multi-rider detection.

[0005] This invention provides a method for detecting multiple riders riding bicycles, comprising: In response to a user's request to use a target vehicle, the system obtains the user's baseline weight parameter, which is determined based on the user's historical cycling data. Based on the real-time load parameters of the target vehicle and the baseline weight parameters, multi-person riding detection is performed to obtain multi-person riding detection results; the multi-person riding detection results are used to indicate whether there is currently multi-person riding behavior in the target vehicle.

[0006] According to a method for detecting multiple riders provided by the present invention, the historical riding data includes a historical load dataset, which contains load data from multiple rides by the user; obtaining the user's baseline weight parameter includes: Based on the user's user information, a parameter retrieval request is generated and sent to the cloud platform; Receive the user's baseline weight parameters sent by the cloud platform; The baseline weight parameter is a reference value representing the user's weight in a single-person cycling state, obtained by the cloud platform after analyzing the historical load dataset; the baseline weight parameter is dynamically updated.

[0007] According to the present invention, a multi-person cycling detection method is provided, wherein the reference weight parameter is the average value of each load data in the light load data group; The lightweight load data set is a set of load data with smaller values ​​obtained by grouping the historical load dataset when the variance of the historical load dataset is greater than the instability threshold.

[0008] According to the multi-person cycling detection method provided by the present invention, the lightweight load data group is a group of load data with smaller values ​​obtained by grouping the load data sequence with the target load data as the boundary. The load data sequence is a numerical sequence obtained by sorting each load data in the historical load dataset. The target load data is the load data corresponding to the maximum load difference in the load data sequence, and the maximum load difference is the maximum value among the differences between adjacent load data in the load data sequence.

[0009] According to a method for detecting multiple riders provided by the present invention, the step of detecting multiple riders based on the real-time load parameters of the target vehicle and the reference weight parameters to obtain the detection result includes: Based on the real-time load parameters and the baseline weight parameters, the load increment value is determined; If the load increment value is greater than the increment threshold, and / or the ratio of the load increment value to the baseline weight parameter is greater than the proportion threshold, then the multi-person riding detection result is determined to be that the target vehicle currently has multi-person riding behavior. Otherwise, the result of the multi-person riding detection is determined to be that there is currently no multi-person riding behavior on the target vehicle.

[0010] According to a multi-person cycling detection method provided by the present invention, the incremental threshold and / or the proportional threshold are determined based on the type of operating area where the target vehicle is located, and the type of operating area is determined based on the vehicle location coordinates of the target vehicle or the area range in which the target vehicle is allowed to operate.

[0011] The present invention also provides a multi-person cycling detection device, comprising: The acquisition unit is used to acquire the user's baseline weight parameters in response to the user's request to use the target vehicle. The baseline weight parameters are determined based on the user's historical cycling data. The detection unit is used to perform multi-person riding detection based on the real-time load parameters of the target vehicle and the baseline weight parameters, and obtain multi-person riding detection results; the multi-person riding detection results are used to indicate whether there is currently multi-person riding behavior in the target vehicle.

[0012] The present invention also provides a target vehicle, including a weight measuring device and a multi-person riding detection device as described above; The weight measuring device is used to detect the real-time load parameters of the target vehicle.

[0013] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the multi-person cycling detection method as described above.

[0014] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the multi-person cycling detection method as described above.

[0015] The multi-person cycling detection method, device, vehicle, electronic equipment, and storage medium provided by this invention abandon the fixed threshold judgment logic of traditional solutions that apply a one-size-fits-all approach to all users. Instead, it introduces a personalized benchmark weight parameter based on historical cycling data, breaking the limitations of fixed thresholds. It can accurately identify multiple-person cycling based on the historical weight characteristics of each user. This not only effectively solves the problem of heavy users being easily misjudged as riding as a duo and thus unable to use the vehicle, but also solves the problem of missed detection when two light users ride at the same time because their total weight does not exceed the fixed threshold. Even when two people get on the vehicle at the same time and there is no sudden change in weight, it can accurately identify multi-person cycling behavior by comparing the huge difference between the total weight and the benchmark weight parameter. Thus, it greatly improves the accuracy and robustness of detection without significantly increasing hardware costs. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating the multi-person cycling detection method provided by the present invention; Figure 2 This is a schematic diagram of the multi-person cycling detection device provided by the present invention; Figure 3This is a schematic diagram of the structure of the target vehicle provided by the present invention; Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0019] With the rapid development of the sharing economy, shared electric bicycles, as a green and convenient short-distance travel tool, have become deeply integrated into people's daily lives. However, the phenomenon of multiple riders persists despite repeated bans. This behavior not only seriously affects the vehicle's handling balance, increases braking distance, and raises the risk of traffic accidents, but also violates relevant traffic regulations, posing significant hidden dangers to user safety and business operations. Therefore, how to accurately detect and stop multiple riders has become an urgent problem to be solved in this field.

[0020] Currently, the main method is to install gravity sensors on the vehicle to detect changes in load, thereby determining whether multiple riders are involved. For example, this can be achieved by detecting the maximum change in load data within a preset time period, or by directly determining whether the current load exceeds a preset maximum load.

[0021] This method is effective in scenarios where multiple people take turns boarding. For example, if the first user boards and generates a 45kg load, and then the second user boards and generates a 45kg change in load, this change can be detected as overloading. However, in practical applications, this detection logic based on a fixed threshold or a simple change in load presents an irreconcilable contradiction, especially when facing the following complex scenarios, where the shortcomings are particularly evident: On the one hand, there is a risk of underreporting. For example, if two lighter users, such as two women weighing 45kg each, get on the vehicle at the same time, the total weight the vehicle can bear will instantly change from 0 to 90kg. Because the two people move in sync, the sensors cannot detect the change in weight gain. If the maximum load data is set to 90kg to avoid false alarms, this type of two-person riding behavior will not be recognized, leading to a security vulnerability.

[0022] On the other hand, there is a risk of misjudgment. To address the aforementioned underreporting issue, simply lowering the maximum load data, such as setting it to less than 90kg, would result in a situation where a heavier adult male, such as 120kg, scans the code to use the bike, and the detected initial load exceeds the maximum load data. This would lead to the bike being incorrectly classified as being ridden by multiple people or overloaded, preventing the legitimate user from using the bike and severely impacting the user experience.

[0023] In summary, due to the significant difference between the lower limit of adult female weight and the upper limit of adult male weight, and the fact that the weight of a single rider is often greater than the sum of the weights of two riders, the existing technology, which relies on general and fixed thresholds, cannot simultaneously meet the needs of "detecting two light-weight users getting on the bike at the same time" and "ensuring normal use of the bike by a heavy-weight user." This has become an irreconcilable contradiction in the existing technology.

[0024] In response, this invention provides a method for detecting multiple riders, which aims to introduce a baseline weight parameter based on the user's historical riding data to break through the limitations of traditional fixed threshold detection and achieve accurate identification of multiple rider behavior. It can not only effectively detect violations such as multiple users boarding the bike at the same time or taking turns to board, but also accurately distinguish between heavy riders riding alone and light riders riding in groups, thereby avoiding misjudgment of heavy users and greatly optimizing the user experience while ensuring operational safety.

[0025] Figure 1 This is a flowchart illustrating the multi-person riding detection method provided by the present invention. This method is applied in the operation and management of shared vehicles, and the specific executing entity can be the controller on the target vehicle, such as an in-vehicle central control controller. Figure 1 As shown, the method includes: Step 110: In response to the user's request to use the target vehicle, obtain the user's baseline weight parameter, which is determined based on the user's historical cycling data; Step 120: Perform multi-person riding detection based on the real-time load parameters of the target vehicle and the baseline weight parameters to obtain the multi-person riding detection results; the multi-person riding detection results are used to indicate whether there is currently multi-person riding behavior in the target vehicle.

[0026] Specifically, when a user wishes to use a target vehicle, they may initiate a vehicle usage request. This request can be an unlocking command triggered by the user scanning a QR code on the vehicle's body via an application on their mobile terminal; it can also be a usage command generated through near-field communication touch, Bluetooth connection pairing, or manually entering the vehicle's serial number; or it can be other forms of commands that characterize the user's usage needs. This embodiment of the invention does not specifically limit these. Here, the target vehicle has load detection and network communication capabilities, and can be a shared electric vehicle, shared bicycle, electric scooter, etc.

[0027] Upon receiving a user's request to use the vehicle, this embodiment of the invention does not simply perform an unlocking action as in traditional solutions. Instead, it acquires crucial data, specifically the user's baseline weight parameter. This baseline weight parameter is not a universal, fixed value, but rather a personalized value representing the user's normal weight under single-person riding conditions.

[0028] It is worth noting that this baseline weight parameter is determined based on the user's historical cycling data. This means that it is necessary to pre-record and store the user's historical cycling data generated in one or more past rides. By conducting in-depth analysis, statistics, and filtering of this historical cycling data (such as averaging after removing outliers, weighted calculation, or extracting the median), a user profile can be created, depicting the user's true individual weight from a data perspective, thereby determining the user's baseline weight parameter.

[0029] In practical applications, for example, when a user scans a code to request a ride, the target vehicle will upload the user ID (IdentityDocument) to the cloud platform. After receiving the user ID, the cloud platform will retrieve the historical profile corresponding to the user ID from the database, extract the pre-calculated baseline weight parameters, for example, the user's usual single weight is 75kg, and send the parameter to the target vehicle or temporarily store it in the cloud for later use.

[0030] When a user starts using the vehicle, the target vehicle continuously or periodically collects real-time load parameters through its onboard weight measurement devices, such as gravity sensors or pressure sensors installed in the seat, axle, or bracket. These parameters directly reflect the actual total weight acting on the target vehicle at the current moment (or voltage or pressure physical quantities that are mapped to weight).

[0031] After this, the collected real-time load parameters can be logically calculated and compared with the user's baseline weight parameters to achieve multi-person cycling detection. The core of this detection process is to determine whether there is an unreasonable deviation between the current actual weight and the user's usual baseline weight. Specifically, the detection logic here can be: calculating the difference between the real-time load parameters and the baseline weight parameters (i.e., the load increment value).

[0032] Furthermore, if the difference is within a preset reasonable fluctuation range, for example, if the difference is very small, or only equivalent to the weight of a backpack, then it can be considered that the current activity is still a solo ride.

[0033] Conversely, if the difference exceeds the reasonable fluctuation range, for example, if it reaches more than 40kg or 50% of the baseline weight, it means that the load on the target vehicle has changed abnormally and exceeded the fluctuation range of the user's daily weight. It is very likely that an additional passenger has been added, and at this time it can be considered that multiple people are riding.

[0034] After the above comparison and analysis, the multi-person riding detection result can be obtained. This result is used to clearly indicate whether there is multi-person riding behavior in the target vehicle. Its form can be a Boolean command used to control the vehicle alarm or power off, such as "yes / no", or it can be a specific load status code. This embodiment of the invention does not make specific limitations on this.

[0035] The multi-person cycling detection method provided by this invention abandons the fixed threshold judgment logic of traditional solutions that apply a one-size-fits-all approach to all users. Instead, it introduces a personalized baseline weight parameter based on historical cycling data, breaking the limitations of fixed thresholds. It can accurately identify multiple-person cycling based on the historical weight characteristics of each user. This not only effectively solves the problem of heavy users being easily misjudged as riding as a duo and thus unable to use the bike, but also solves the problem of missed detection when two light users ride at the same time because their total weight does not exceed the fixed threshold. Even when two people get on the bike at the same time and there is no sudden change in weight, it can accurately identify multi-person cycling behavior by comparing the huge difference between the total weight and the baseline weight parameter. Thus, it greatly improves the accuracy and robustness of detection without significantly increasing hardware costs.

[0036] Based on the above embodiments, historical cycling data includes a historical load dataset, which contains load data from multiple rides by the user. In step 110, the user's baseline weight parameters are obtained, including: Based on the user's user information, a parameter retrieval request is generated and sent to the cloud platform; Receive the user's baseline weight parameters from the cloud platform; The baseline weight parameter is a reference value representing the user's weight in a single-person cycling state, obtained by the cloud platform based on the analysis of historical load datasets; the baseline weight parameter is dynamically updated.

[0037] In actual operation, considering the limitations of storage space and computing power of the target vehicle, as well as the need for cross-vehicle sharing of user data, this embodiment of the invention adopts an edge-cloud collaborative processing architecture.

[0038] Specifically, the process of obtaining the user's baseline weight parameters includes: When a user's vehicle request is received, it's necessary to determine who is currently using the vehicle. Therefore, the user's information is extracted, such as user ID, registered mobile phone number, and encrypted identity token. A parameter retrieval request is then generated based on this user information and sent to the cloud platform. This parameter retrieval request is an instruction packet containing user information, designed to request the user's baseline weight parameters from the cloud platform, which possesses powerful data processing capabilities.

[0039] After receiving the parameter acquisition request, the cloud platform will transmit the calculated baseline weight parameters back to the target vehicle via wireless networks such as 4G (The 4th Generation Mobile Communication Technology), 5G (5th-Generation Mobile Communication Technology), and NB-IoT (Narrow Band Internet of Things).

[0040] To provide accurate baseline weight parameters, the cloud platform maintains the user's historical cycling data. In this embodiment, historical cycling data includes a historical load dataset, which contains load data from multiple rides by the user. This means that the cloud platform not only records the user's most recent ride, but also, like a diary, accumulates load data from multiple rides taken by the user at different times and using different vehicles. This aggregated data constitutes a historical load dataset that reflects the user's long-term weight change patterns.

[0041] The cloud platform leverages its powerful computing capabilities to analyze this massive historical load dataset. The cloud-based analysis logic extracts a net weight value that purely reflects the user's own weight from the historical load dataset, which includes various possible scenarios such as single-person riding, multi-person riding, and riding with cargo. For example, statistical methods are used to remove abnormally high data (potentially indicating carrying passengers) or abnormally low data (potentially false readings), thus identifying the data that best represents the user's physical characteristics as the baseline weight parameter.

[0042] More importantly, the baseline weight parameter in this embodiment of the invention is not fixed, but changes continuously over time with the accumulation of cycling sessions. That is, the baseline weight parameter is dynamically updated.

[0043] Specifically, a user's actual weight is not static. Factors such as seasonal changes in clothing and variations in the user's own weight can influence this. Furthermore, as the number of rides increases, the cloud platform collects more historical riding data, leading to a more accurate user profile. Therefore, dynamic updates mean that the cloud platform continuously adjusts this baseline weight parameter based on newly generated riding data. Each time a user uses the bike, the baseline weight parameter obtained is calculated based on the latest and most comprehensive historical load dataset, best reflecting the user's current condition, rather than the data entered during registration.

[0044] In this embodiment of the invention, on the one hand, by utilizing an edge-cloud collaborative architecture, the complex tasks of storing and analyzing historical cycling data are transferred to the cloud platform for execution, greatly reducing the hardware burden on the target vehicle and enabling seamless roaming of user characteristics between different vehicles; on the other hand, by establishing a historical load dataset containing load data from multiple rides and dynamically updating it, the baseline weight parameters have adaptive capabilities, and the detection process always uses the latest reference values ​​for comparison, thereby maintaining a very high accuracy rate in detecting multiple riders over a long operating cycle and avoiding misjudgments caused by outdated data.

[0045] Based on the above embodiments, the baseline weight parameter is the average value of each load data in the lightweight load data group; Lightweight load data sets are groups of load data with smaller values ​​obtained by grouping historical load datasets when the variance of the historical load dataset exceeds the instability threshold.

[0046] Specifically, in the historical load dataset maintained by the cloud platform, the data should ideally be relatively clustered and stable. However, in actual operation, some users may have intermittent multi-person riding behavior in their past riding records, or occasionally carry heavy objects, resulting in their historical load dataset being mixed with load data from various states such as normal single-person riding and multi-person / loaded riding, forming interfering dirty data.

[0047] To extract a user's true individual weight from this data, the cloud platform first assesses the dispersion of the dataset before calculation. Specifically, it calculates the variance of the historical load dataset and compares this variance to a preset instability threshold. This instability threshold is an empirical value used to define whether data fluctuations are within an acceptable range.

[0048] When the variance of the historical load dataset exceeds the instability threshold, it indicates that the user's load data fluctuates drastically. For example, the load data may jump between 50kg and 110kg. In this case, the historical load dataset can be determined to be unstable, and it is highly likely that it contains data from different riding conditions. If all load data are simply averaged in this situation, the calculated baseline weight parameter will be artificially inflated by abnormally high values, leading to overly lenient subsequent detection standards.

[0049] To address this scenario, this embodiment of the invention groups the historical load dataset. This process can utilize clustering algorithms or numerical breakpoint analysis to decompose the discrete historical load dataset into several different data clusters. Based on common sense, the total load of multiple riders is significantly greater than that of a single rider. Therefore, after grouping, the group with the smaller load data is automatically identified and locked, and marked as the lightweight load data group. This data group is considered to represent the actual load of a user during a standard single-rider ride.

[0050] Subsequently, the data set with the larger values ​​will be discarded, and only the selected lightweight load data set will be used for calculation. That is, the user's baseline weight parameter will be determined by calculating the arithmetic mean of the lightweight load data set.

[0051] In this embodiment of the invention, the introduction of variance determination and data grouping strategies greatly enhances the anti-interference capability of the detection process. Even if a user has repeatedly violated regulations by carrying passengers in their historical behavior, resulting in a large number of high-load records in the historical load data set, this situation can be detected by variance anomalies. The high-load data can be intelligently separated using a grouping and filtering mechanism, and the light-load data set belonging to single-person riding can be accurately purified. This ensures that the generated benchmark weight parameters are always anchored to the user's actual weight level, effectively preventing the risk of the judgment standard being maliciously raised due to the contamination of historical violation data.

[0052] Based on the above embodiments, the lightweight load data group is a group of load data with smaller values ​​obtained by grouping the load data sequence with the target load data as the boundary. The load data sequence is a numerical sequence obtained by sorting the load data in the historical load dataset. The target load data is the load data corresponding to the maximum load difference in the load data sequence. The maximum load difference is the maximum value among the differences between adjacent load data in the load data sequence.

[0053] Specifically, after determining that the historical load dataset, which exhibits drastic fluctuations and excessive variance, needs to be grouped, this embodiment of the invention first performs data ordering processing. That is, the load data in the historical load dataset is arranged in ascending order to obtain a load data sequence. This allows the originally chaotic load data to exhibit clustering on the numerical axis, causing load data with similar values ​​to be grouped together.

[0054] Next, we can find the dividing line between single-person cycling data and multi-person cycling data in the load data sequence. Specifically, this can be done by iterating through the load data sequence and calculating the difference between any two adjacent load data points. The largest difference among all calculated differences is defined as the maximum load difference.

[0055] It's important to note here that data from the same type of behavior (such as all single-person cycling) tends to be relatively compact, with small differences between adjacent values. However, between different types of behavior (such as a jump from single-person cycling to multi-person cycling), there must be a significant weight difference. Therefore, the breakpoint where the maximum load difference—that is, the maximum value among the differences between adjacent load data in the load data sequence—is located is precisely the optimal dividing point for distinguishing between single-person cycling and multi-person cycling.

[0056] After that, the load data at the gap position (specifically, the side before the gap, i.e. the side with the smaller value) can be marked as the target load data, and the data can be grouped with this as the boundary. The subsequence containing the target load data itself and all smaller values ​​before it is the light load data group; while the large value group after the gap is regarded as multi-person riding data and is removed.

[0057] The following is a specific example to illustrate the data grouping process: Suppose that the historical load data set of a user obtained by the cloud platform is [75, 40, 80, 42, 71, 49] (unit: kg).

[0058] By sorting the dataset in ascending order, the load data sequence [40, 42, 49, 71, 75, 80] is obtained.

[0059] Calculate the differences between adjacent load data in the load data sequence, which are 2, 7, 22, 4, and 5 respectively.

[0060] The maximum load difference was determined, with the largest jump (22 kg) occurring between values ​​49 and 71. Value 49, located to the left of the jump point (the smaller side), was identified as the target load data.

[0061] The load data sequence was divided into two groups, [40, 42, 49] and [71, 75, 80], with 49 as the boundary. The first group, [40, 42, 49], with the smaller value, was selected as the lightweight load data group.

[0062] Calculate the average value of the light load data set, which is approximately (40+42+49) / 3≈43.6kg. Use this average value as the user's baseline weight parameter.

[0063] In this embodiment of the invention, the huge physical gap between the weight of a single person and the weight of multiple people in the numerical space is utilized. Without relying on complex machine learning models, the data distribution gaps can be quickly and accurately located through simple sorting and difference calculation. This maximum difference classification method has extremely high robustness and can automatically adapt to the weight base of different users, accurately separating the mixed data. This ensures that the extracted lightweight load data group has extremely high purity, restores the user's true single weight to the greatest extent, and ensures the accuracy of detection.

[0064] Based on the above embodiments, step 120 includes: The load increment value is determined based on real-time load parameters and baseline weight parameters; If the load increment is greater than the increment threshold, and / or the ratio of the load increment to the baseline weight parameter is greater than the proportion threshold, then the multi-person riding detection result is determined to indicate that the target vehicle currently has multi-person riding behavior; otherwise, the multi-person riding detection result is determined to indicate that the target vehicle currently does not have multi-person riding behavior.

[0065] Specifically, after obtaining the real-time load parameters of the target vehicle and the user's baseline weight parameters, the core task of detection becomes quantifying the degree of difference between the two. Based on this, in this embodiment of the invention, the load increment is calculated according to the real-time load parameters and the baseline weight parameters to obtain the load increment value. This load increment value intuitively reflects the increased weight in the current riding state compared to the user's usual solo riding state. To determine whether this weight belongs to normal personal belongings (such as backpacks or shopping bags) or to passengers violating regulations, this embodiment of the invention introduces a dual judgment standard consisting of an increment threshold and a percentage threshold. Here, the increment threshold is a preset weight value, for example, set to 35kg or 40kg; the percentage threshold is a preset percentage coefficient, for example, set to 40% or 50%.

[0066] In detail, if the load increment value is greater than the increment threshold, it means that the added weight is objectively large, and is very likely the weight of a person; and / or, if the ratio of the load increment value to the baseline weight parameter is greater than the proportion threshold, it means that the load on the vehicle has increased unreasonably and drastically relative to the user's own physique. For example, a 40kg user suddenly gains 25kg of weight. Although it does not exceed the absolute threshold, the proportion is too high, and it is very likely that the user is carrying a child. At this time, it can be determined that the multi-person riding detection result indicates that there is currently multi-person riding behavior on the target vehicle, and the corresponding control signal is output.

[0067] Correspondingly, if the increase in load does not exceed the increment threshold and its relative proportion is within a reasonable proportion threshold range, for example, if an 80kg user carries a 10kg bag, the increase is 10kg, accounting for 12.5%, then the weight fluctuation can be determined to be a normal deviation for single-person riding. In this case, the detection result for multi-person riding can be determined to be that there is currently no multi-person riding behavior.

[0068] In this embodiment of the invention, a rigorous and inclusive detection logic is constructed by combining an absolute incremental threshold with a relative proportional threshold. The incremental threshold directly intercepts obvious adult-carrying behavior; while the proportional threshold effectively addresses the detection shortcomings for lighter users carrying passengers, especially children. This dual-insurance mechanism accurately distinguishes between carrying people and carrying objects while ensuring fairness in detection for users of different weights, significantly improving the accuracy and reliability of the detection results.

[0069] Based on the above embodiments, the incremental threshold and / or proportional threshold are determined based on the type of operating area where the target vehicle is located, and the type of operating area is determined based on the vehicle location coordinates of the target vehicle or the area range in which the target vehicle is allowed to operate.

[0070] Specifically, considering that in the actual urban operation network of shared vehicles, different geographical blocks often gather different types of user groups, and these groups have significantly different travel and carrying habits, in order to balance the accuracy of detection and user experience, the judgment criteria set in this embodiment of the invention, namely the incremental threshold and / or the proportional threshold, are not unchanging global constants, but variables determined based on the type of operating area where the target vehicle is located.

[0071] To implement this mechanism, in this embodiment of the invention, it is first necessary to sense the vehicle's location. This can be achieved by using the vehicle's built-in positioning system, such as GPS (Global Positioning System) or BeiDou positioning modules, to collect the target vehicle's location coordinates. Alternatively, the approximate location of the target vehicle can be determined by judging whether it is currently within the area permitted for operation by the cloud platform, i.e., an electronic fence block.

[0072] Next, this location information can be matched with the map semantic database in the background to identify the type of operating area where the target vehicle is currently located. For example, university campuses, residential communities, CBD (Central Business District) business districts, subway connection points, etc.

[0073] For different identified operating area types, this embodiment of the invention will automatically issue or switch appropriate threshold parameters. Specifically: If the target vehicle is identified as being located on a university campus, considering the common need for students to ride with heavy backpacks (including textbooks, laptops, etc.), setting the threshold too low could easily lead to misjudgments. Therefore, in this case, the incremental threshold will be automatically increased. For example, the standard value of 35kg will be relaxed to 45kg, giving the area a higher load tolerance and preventing students with backpacks from being mistaken for multiple riders.

[0074] If the target vehicle is identified as being in a residential community, given the higher risk of riding with children and the children's lighter weight, the proportional threshold may be lowered and / or a strict standard incremental threshold may be maintained to keep a high level of sensitivity to violations of passenger carrying regulations.

[0075] By using the above methods, the incremental threshold and / or proportional threshold can be dynamically tightened or relaxed according to changes in the geographical scene.

[0076] In this embodiment of the invention, a scenario-based threshold adjustment mechanism based on geofencing effectively solves the problem of false alarms caused by carrying heavy objects normally in special scenarios, such as school students carrying backpacks. While ensuring strict adherence to safety standards, it minimizes disturbance to normal users, greatly optimizes the user experience, achieves refined operation management, and significantly improves the environmental adaptability of the detection method.

[0077] The multi-person cycling detection device provided by the present invention is described below. The multi-person cycling detection device described below can be referred to in correspondence with the multi-person cycling detection method described above.

[0078] Figure 2 This is a structural schematic diagram of the multi-person cycling detection device provided by the present invention, as shown below. Figure 2 As shown, the device includes: The acquisition unit 210 is used to acquire the user's baseline weight parameter in response to the user's request to use the target vehicle. The baseline weight parameter is determined based on the user's historical cycling data. The detection unit 220 is used to perform multi-person riding detection based on the real-time load parameters of the target vehicle and the baseline weight parameters, and obtain multi-person riding detection results; the multi-person riding detection results are used to indicate whether there is currently multi-person riding behavior in the target vehicle.

[0079] The multi-person cycling detection device provided by this invention abandons the fixed threshold judgment logic of traditional solutions that apply a one-size-fits-all approach to all users. Instead, it introduces a personalized benchmark weight parameter based on historical cycling data, breaking the limitations of fixed thresholds. It can accurately identify multiple-person cycling based on the historical weight characteristics of each user. This not only effectively solves the problem of heavy users being easily misjudged as riding as a duo and thus unable to use the bike, but also solves the problem of missed detection when two light users ride at the same time because their total weight does not exceed the fixed threshold. Even when two people get on the bike at the same time and there is no sudden change in weight, it can accurately identify multi-person cycling behavior by comparing the huge difference between the total weight and the benchmark weight parameter. Thus, it greatly improves the accuracy and robustness of detection without significantly increasing hardware costs.

[0080] Based on the above embodiments, the historical cycling data includes a historical load dataset, which contains load data from multiple rides by the user. The acquisition unit 210 is used for: Based on the user's user information, a parameter retrieval request is generated and sent to the cloud platform; Receive the user's baseline weight parameters sent by the cloud platform; The baseline weight parameter is a reference value representing the user's weight in a single-person cycling state, obtained by the cloud platform after analyzing the historical load dataset; the baseline weight parameter is dynamically updated.

[0081] Based on the above embodiments, the baseline weight parameter is the average value of each load data in the lightweight load data group; The lightweight load data set is a set of load data with smaller values ​​obtained by grouping the historical load dataset when the variance of the historical load dataset is greater than the instability threshold.

[0082] Based on the above embodiments, the lightweight load data group is a group of load data with smaller values ​​obtained by grouping the load data sequence with the target load data as the boundary. The load data sequence is a numerical sequence obtained by sorting each load data in the historical load dataset. The target load data is the load data corresponding to the maximum load difference in the load data sequence, and the maximum load difference is the maximum value among the differences between adjacent load data in the load data sequence.

[0083] Based on the above embodiments, the detection unit 220 is used for: Based on the real-time load parameters and the baseline weight parameters, the load increment value is determined; If the load increment value is greater than the increment threshold, and / or the ratio of the load increment value to the baseline weight parameter is greater than the proportion threshold, then the multi-person riding detection result is determined to be that the target vehicle currently has multi-person riding behavior. Otherwise, the result of the multi-person riding detection is determined to be that there is currently no multi-person riding behavior on the target vehicle.

[0084] Based on the above embodiments, the incremental threshold and / or the proportional threshold are determined based on the type of operating area where the target vehicle is located, and the type of operating area is determined based on the vehicle location coordinates of the target vehicle or the area range in which the target vehicle is allowed to operate.

[0085] The present invention also provides a target vehicle, Figure 3 This is a structural schematic diagram of the target vehicle provided by the present invention, such as... Figure 3 As shown, the vehicle includes a weight measuring device 310 and a multi-person riding detection device 320 as described above. The weight measuring device 310 is used to detect the real-time load parameters of the target vehicle.

[0086] Figure 4 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 4 As shown, the electronic device may include a processor 410, a communications interface 420, a memory 430, and a communication bus 440, wherein the processor 410, communications interface 420, and memory 430 communicate with each other via the communication bus 440. The processor 410 can call logical instructions in the memory 430 to execute a multi-person riding detection method. This method includes: in response to a user's request to use the target vehicle, obtaining the user's baseline weight parameter, which is determined based on the user's historical riding data; performing multi-person riding detection based on the target vehicle's real-time load parameter and the baseline weight parameter to obtain a multi-person riding detection result; the multi-person riding detection result is used to indicate whether multi-person riding behavior currently exists in the target vehicle.

[0087] Furthermore, the logical instructions in the aforementioned memory 430 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0088] On the other hand, the present invention also provides a computer program product, the computer program product comprising a computer program stored on a non-transitory computer-readable storage medium, the computer program comprising program instructions, wherein when the program instructions are executed by a computer, the computer is able to execute the multi-person riding detection method provided by the above methods, the method comprising: in response to a user's request to use a target vehicle, obtaining the user's baseline weight parameter, the baseline weight parameter being determined based on the user's historical riding data; performing multi-person riding detection based on the target vehicle's real-time load parameter and the baseline weight parameter, and obtaining a multi-person riding detection result; the multi-person riding detection result being used to indicate whether there is currently multi-person riding behavior in the target vehicle.

[0089] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the multi-person cycling detection method provided by the methods described above. The method includes: in response to a user's request to use a target vehicle, obtaining a baseline weight parameter of the user, the baseline weight parameter being determined based on the user's historical cycling data; performing multi-person cycling detection based on the real-time load parameter of the target vehicle and the baseline weight parameter, obtaining a multi-person cycling detection result; the multi-person cycling detection result being used to indicate whether multi-person cycling behavior currently exists in the target vehicle.

[0090] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0091] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0092] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for detecting multiple riders, characterized in that, include: In response to a user's request to use a target vehicle, the system obtains the user's baseline weight parameter, which is determined based on the user's historical cycling data. Based on the real-time load parameters of the target vehicle and the baseline weight parameters, a multi-person riding detection is performed to obtain the multi-person riding detection results. The multi-person riding detection result is used to indicate whether there is currently multi-person riding behavior in the target vehicle.

2. The multi-person cycling detection method according to claim 1, characterized in that, The historical cycling data includes a historical load dataset, which contains load data from multiple rides by the user. The process of obtaining the user's baseline weight parameters includes: Based on the user's user information, a parameter retrieval request is generated and sent to the cloud platform; Receive the user's baseline weight parameters sent by the cloud platform; The baseline weight parameter is a reference value representing the user's weight in a single-person cycling state, obtained by the cloud platform after analyzing the historical load dataset; the baseline weight parameter is dynamically updated.

3. The multi-person cycling detection method according to claim 2, characterized in that, The baseline weight parameter is the average value of each load data in the lightweight load data group; The lightweight load data set is a set of load data with smaller values ​​obtained by grouping the historical load dataset when the variance of the historical load dataset is greater than the instability threshold.

4. The multi-person cycling detection method according to claim 3, characterized in that, The lightweight load data group is a group of load data with smaller values ​​obtained by grouping the load data sequence with the target load data as the boundary. The load data sequence is a numerical sequence obtained by sorting each load data in the historical load dataset. The target load data is the load data corresponding to the maximum load difference in the load data sequence, and the maximum load difference is the maximum value among the differences between adjacent load data in the load data sequence.

5. The method for detecting multiple riders according to any one of claims 1 to 4, characterized in that, The multi-person riding detection based on the real-time load parameters of the target vehicle and the baseline weight parameters, to obtain the multi-person riding detection results, includes: Based on the real-time load parameters and the baseline weight parameters, the load increment value is determined; If the load increment value is greater than the increment threshold, and / or the ratio of the load increment value to the baseline weight parameter is greater than the proportion threshold, then the multi-person riding detection result is determined to be that the target vehicle currently has multi-person riding behavior. Otherwise, the result of the multi-person riding detection is determined to be that there is currently no multi-person riding behavior on the target vehicle.

6. The multi-person cycling detection method according to claim 5, characterized in that, The incremental threshold and / or the proportional threshold are determined based on the type of operating area where the target vehicle is located, and the type of operating area is determined based on the vehicle location coordinates of the target vehicle or the area range in which the target vehicle is allowed to operate.

7. A multi-person cycling detection device, characterized in that, include: The acquisition unit is used to acquire the user's baseline weight parameters in response to the user's request to use the target vehicle. The baseline weight parameters are determined based on the user's historical cycling data. The detection unit is used to perform multi-person riding detection based on the real-time load parameters of the target vehicle and the reference weight parameters, and obtain the multi-person riding detection results. The multi-person riding detection result is used to indicate whether there is currently multi-person riding behavior in the target vehicle.

8. A target vehicle, characterized in that, Includes a weight measuring device and a multi-person cycling detection device as described in claim 7; The weight measuring device is used to detect the real-time load parameters of the target vehicle.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the multi-person cycling detection method as described in any one of claims 1 to 6.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the multi-person cycling detection method as described in any one of claims 1 to 6.

Images