A monitoring and early warning method and system for a taxi

Abstract

The application discloses a taxi monitoring and early warning method and system, and relates to the technical field of artificial intelligence.The method comprises the following steps: collecting historical taxi operation data, and constructing model training samples based on the historical taxi operation data; training a taxi monitoring and early warning model based on the model training samples to obtain a trained taxi monitoring and early warning model; transmitting the mileage charging data of a taxi meter by using encrypted pulse signals, and judging whether the frequency of the encrypted pulse signals is abnormal in real time through the self-checking function of the taxi meter; if the frequency of the encrypted pulse signals is abnormal, generating signal abnormality early warning information; collecting real-time taxi operation data, inputting the real-time taxi operation data and the signal abnormality early warning information into the trained taxi monitoring and early warning model, and obtaining a taxi operation abnormal behavior judgment result, which helps to solve the problem that the prior art cannot realize real-time, accurate and full-process information monitoring and early warning on the illegal installation of a small motor cheating behavior of a taxi.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and in particular to a method and system for monitoring and early warning of taxis. Background Technology

[0002] The "small engine" monitoring model for illegal taxi operations is based on traffic big data, incorporating traffic enforcement and supervision data, order data, location data, and road condition data to achieve "multi-source data convergence and centralized empowerment," and realizing digital supervision of taxis through big data and artificial intelligence technologies. However, this type of monitoring method still has many technical shortcomings in practical applications, making it difficult to meet the needs of refined and intelligent supervision: First, the information flow efficiency is low, relying on manual reporting, which prevents higher-level regulatory departments from obtaining operational data in a timely manner, resulting in delayed regulatory response; Second, data sharing is hampered, with no vertical data exchange between traffic enforcement and supervision departments in different regions, and a lack of horizontal cross-departmental data sharing, resulting in incomplete regulatory information; Third, data analysis methods are limited, with insufficient mining and utilization of multi-source operational data, making it difficult to accurately identify cheating behavior; Fourth, there is a lack of information-based supervision methods, still relying mainly on traditional static supervision, failing to achieve the integration of static management and dynamic monitoring; Fifth, meter signals are easily tampered with, the authenticity of mileage billing data cannot be guaranteed, and the difficulty in obtaining evidence of cheating behavior further reduces the effectiveness of supervision.

[0003] Therefore, there is an urgent need for a method that can achieve real-time, accurate, and full-process information-based monitoring and early warning of illegal installation of small motors in taxis for cheating. Summary of the Invention

[0004] In view of this, the present invention proposes a monitoring and early warning method and system for taxis, which can realize real-time, accurate, and full-process information-based monitoring and early warning of taxis illegally installing small motors to cheat.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A method for monitoring and early warning of taxis, comprising: Collect historical taxi operation data, and construct model training samples based on the historical taxi operation data; The taxi monitoring and early warning model is trained based on the training samples of the model to obtain a trained taxi monitoring and early warning model; The taxi meter uses encrypted pulse signals to transmit mileage fare data. The built-in self-test function of the taxi meter can determine in real time whether the frequency of the encrypted pulse signal is abnormal. If so, a signal abnormality warning message is generated. Real-time taxi operation data is collected, and the real-time taxi operation data and the signal anomaly warning information are input into a trained taxi monitoring and warning model to obtain the judgment result of abnormal taxi operation behavior.

[0006] Based on the above technical solution, the present invention can be further improved as follows: Optionally, the historical taxi operation data includes taxi order data, taxi trajectory data, key vehicle database data, credit assessment results data, complaint work order data, urban traffic condition data, and traffic violation data.

[0007] Optionally, after the step of collecting historical taxi operating data, the following steps may also be included: The historical taxi operation data is preprocessed, including aggregation, cleaning, deduplication, fusion, standardization, and format unification.

[0008] Optionally, after the step of collecting historical taxi operating data, the following steps may also be included: A digital profile of people and vehicles is constructed based on the historical taxi operation data.

[0009] Optionally, training the taxi monitoring and early warning model based on the model training samples to obtain a trained taxi monitoring and early warning model includes: The model training samples are divided into a training set and a validation set; The training set is input into the taxi monitoring and early warning model for training; The taxi monitoring and early warning model is used to verify its accuracy and optimize its parameters using the validation set. When the anomaly recognition accuracy of the taxi monitoring and early warning model reaches a preset threshold, the training of the taxi monitoring and early warning model is completed, and a trained taxi monitoring and early warning model is obtained.

[0010] Optionally, the step of using the self-test function built into the taxi meter to determine in real time whether the frequency of the encrypted pulse signal is abnormal, and if so, generating a signal abnormality warning message, includes: The frequency is compared with a preset reasonable frequency range. If any of the following situations are detected: frequency exceeding the reasonable range, frequency mutation caused by external device forging encrypted pulse signals, or abnormal increase in frequency during traffic congestion, it is determined that the encrypted pulse signal frequency is abnormal, and a signal abnormality warning message including the abnormality type, the time of abnormality occurrence, and the corresponding meter number is generated.

[0011] Optionally, the taxi monitoring and early warning model includes an input layer, a rule model layer, an unsupervised model layer, a historical behavior comparison layer, a model comprehensive judgment layer, and an output layer; The real-time taxi operation data and the signal anomaly warning information are received through the input layer; By pre-setting thresholds and business logic rules in the rule model layer, hard billing anomalies in taxi operations can be captured; Identify unknown abnormal behaviors that deviate from normal patterns in taxi operations through unsupervised model layers; By comparing the driver's historical billing mileage with the actual mileage on the same road segment based on the digital profile of the driver and vehicle through the historical behavior comparison layer, high-risk operating objects are marked. The model integrates the judgment results of the hard billing anomaly behavior, unknown anomaly behavior and high-risk operation objects through the comprehensive judgment layer, and combines them with the signal anomaly warning information to conduct a comprehensive anomaly analysis; The output layer outputs the results of the judgment on abnormal taxi operation.

[0012] A taxi monitoring and early warning system includes: The data acquisition module is used to collect historical taxi operation data and build model training samples based on the historical taxi operation data. The model training module is used to train the taxi monitoring and early warning model based on the model training samples to obtain a trained taxi monitoring and early warning model. The taxi meter detection module is used to transmit the mileage fare data of the taxi meter using encrypted pulse signals. The self-test function built into the taxi meter is used to determine in real time whether the frequency of the encrypted pulse signal is abnormal. If so, a signal abnormality warning message is generated. The anomaly detection module is used to collect real-time taxi operation data, input the real-time taxi operation data and the signal anomaly warning information into the trained taxi monitoring and warning model, and obtain the judgment result of the taxi's abnormal operation behavior.

[0013] An electronic device includes 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 steps of the method described herein.

[0014] A non-transitory computer-readable storage medium having a computer program stored thereon, the computer program implementing the steps of the method when executed by a processor.

[0015] The present invention has the following advantages: This invention discloses a taxi monitoring and early warning method. It constructs samples from historical operating data and trains a taxi monitoring and early warning model. This model enables intelligent analysis of taxi operational anomalies, improving the accuracy of identifying fraudulent behavior. Simultaneously, it employs encrypted pulse signals to transmit meter fare data, coupled with a meter self-checking function to prevent signal tampering at the source, and promptly generates anomaly warning information. Real-time taxi operating data and signal anomaly warning information are input into the model for comprehensive judgment, achieving end-to-end monitoring from signal tamper prevention to anomaly behavior identification. This effectively solves the problems of distorted fare data and delayed detection of fraudulent behavior in traditional supervision, significantly improving the intelligence and real-time level of taxi operation supervision and providing data support for precise law enforcement. Attached Figure Description

[0016] For illustrative and not limiting purposes, the present invention will now be described in conjunction with embodiments and accompanying drawings, wherein: Figure 1 This is a flowchart illustrating the taxi monitoring and early warning method according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the main components of the taxi monitoring and early warning system according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the first page of the taxi monitoring and early warning system according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the second page of the taxi monitoring and early warning system according to an embodiment of the present invention; Figure 5 This is a schematic diagram of the third page of the taxi monitoring and early warning system according to an embodiment of the present invention; Figure 6 This is a schematic diagram of the fourth page of the taxi monitoring and early warning system according to an embodiment of the present invention; Figure 7 This is a schematic diagram of the fifth page of the taxi monitoring and early warning system according to an embodiment of the present invention; Figure 8 This is a schematic diagram of the physical structure of the electronic device provided by the present invention. Detailed Implementation

[0017] To enable those skilled in the art to better understand the present invention, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0018] It should be noted that the terms "first," "second," etc., in the specification and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be used interchangeably where appropriate for the embodiments of the invention described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0019] It should be noted that, where there is no conflict, the embodiments and features of the present invention can be combined with each other. The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0020] Figure 1 This is a flowchart illustrating the taxi monitoring and early warning method according to an embodiment of the present invention, as shown below. Figure 1 As shown, the taxi monitoring and early warning method provided in this embodiment of the invention includes the following steps S101 to S104.

[0021] S101: Collect historical taxi operation data and build model training samples based on the historical taxi operation data.

[0022] The historical taxi operation data includes taxi order data, taxi trajectory data, key vehicle database data, credit assessment results data, complaint work order data, urban traffic condition data, and traffic violation data.

[0023] Following the step of collecting historical taxi operation data, the following steps are also included: The historical taxi operation data is preprocessed, including aggregation, cleaning, deduplication, fusion, standardization, and format unification, and invalid, redundant, and erroneous data is removed.

[0024] Based on the historical taxi operation data, a digital profile of the driver and vehicle is constructed, recording information such as historical operation characteristics, violation records, and complaint status, providing a basis for subsequent comparison of abnormal behaviors.

[0025] S102, Train the taxi monitoring and early warning model based on the model training samples to obtain the trained taxi monitoring and early warning model.

[0026] The model training samples are divided into a training set and a validation set; The training set is input into the taxi monitoring and early warning model for training; the taxi monitoring and early warning model is a multi-dimensional anomaly detection model designed for taxi cheating behavior, integrating core algorithms such as rule model, unsupervised model, and historical behavior comparison. The taxi monitoring and early warning model is used to verify its accuracy and optimize its parameters using the validation set. When the anomaly recognition accuracy of the taxi monitoring and early warning model reaches a preset threshold, the training of the taxi monitoring and early warning model is completed, and a trained taxi monitoring and early warning model is obtained.

[0027] S103 uses encrypted pulse signals to transmit mileage fare data from the taxi meter. The built-in self-test function of the taxi meter determines in real time whether the frequency of the encrypted pulse signal is abnormal. If so, it generates a signal abnormality warning message.

[0028] First, the core mileage calculation data of the taxi meter is transmitted using encrypted pulse signals, forming an anti-tampering barrier at the source of data transmission. This effectively prevents cheating by external devices that forge or tamper with the mileage pulse signals to modify the mileage data. Simultaneously, the taxi meter's built-in self-checking function is activated. This function collects the frequency data of the encrypted pulse signals transmitted by the meter in real time and compares the collected real-time frequency data with the system's preset reasonable frequency range. If any of the following situations are detected—frequency exceeding the reasonable range, sudden frequency changes caused by external devices forging pulse signals, or abnormal increases in the meter's pulse frequency during periods of traffic congestion when the vehicle is traveling at low speed—it is immediately determined that the encrypted pulse signal frequency is abnormal. Standardized signal anomaly warning information is automatically generated, including the anomaly type, the precise timestamp of the anomaly occurrence, the unique number of the corresponding meter, and the license plate number of the taxi. This provides accurate and traceable clues for subsequent comprehensive anomaly judgment by the model.

[0029] S104: Collect real-time taxi operation data, input the real-time taxi operation data and signal anomaly warning information into the trained taxi monitoring and early warning model, and obtain the judgment result of abnormal taxi operation behavior.

[0030] The real-time taxi operation data and the signal anomaly warning information are received through the input layer; By pre-setting thresholds and business logic rules in the rule model layer, hard billing anomalies in taxi operations can be captured; Identify unknown abnormal behaviors that deviate from normal patterns in taxi operations through unsupervised model layers; By comparing the driver's historical billing mileage with the actual mileage on the same road segment based on the digital profile of the driver and vehicle through the historical behavior comparison layer, high-risk operating objects are marked. The model integrates the judgment results of the hard billing anomaly behavior, unknown anomaly behavior and high-risk operation objects through the comprehensive judgment layer, and combines them with the signal anomaly warning information to conduct a comprehensive anomaly analysis; The output layer outputs the results of the judgment on abnormal taxi operation.

[0031] Specifically, the process is as follows: First, the input layer receives and adapts the data from multiple sources, then synchronously transmits the data to subsequent functional layers. Next, the rule model layer presets mileage anomaly thresholds (e.g., meter mileage exceeding GPS mileage by more than 10%) and business logic rules to quickly capture hard-sounding billing anomalies with clear business logic, such as meter mileage fluctuations when GPS positioning is lost. Then, the unsupervised model layer uses algorithms such as DBSCAN clustering, Isolation Forest / One-ClassSVM, and LSTM-Autoencoder temporal anomaly detection to identify unknown patterns of abnormal behavior, such as taxi spatiotemporal trajectory anomalies, deviations from normal operating records, and vehicle temporal characteristic anomalies. Finally, the historical behavior comparison layer retrieves digital profiles of drivers and vehicles, and compares the drivers' profiles with historical operating data. Historical billing mileage and actual GPS mileage on the same route are compared to identify drivers whose consistently billed mileage exceeds the actual GPS mileage by 5% or more, thus classifying them as having a high-risk operational level. The model's comprehensive judgment layer then performs a multi-dimensional fusion analysis of the hard billing anomaly judgment results from the rule-based model layer, the unknown abnormal behavior judgment results from the unsupervised model layer, and the high-risk operational object marking results from the historical behavior comparison layer. This analysis also incorporates key clues from signal anomaly warning information to conduct a comprehensive and accurate integrated anomaly assessment of taxi operations. Finally, the output layer outputs standardized taxi operational anomaly judgment results, including three levels: normal, suspected anomaly, and confirmed anomaly, along with information such as anomaly type, anomaly severity, and relevant data, providing clear judgment conclusions for precise enforcement by transportation management departments.

[0032] Figure 2 This is a schematic diagram of the main components of the taxi monitoring and early warning system according to an embodiment of the present invention. Figure 2 As shown, the taxi monitoring and early warning system 1 provided in this embodiment of the invention includes a data acquisition module 10, a model training module 20, a fare meter detection module 30, and an anomaly determination module 40.

[0033] Data acquisition module 10 is used to collect historical taxi operation data and construct model training samples based on the historical taxi operation data; Model training module 20 is used to train the taxi monitoring and early warning model based on the model training samples to obtain a trained taxi monitoring and early warning model; The taximeter detection module 30 is used to transmit the mileage fare data of the taxi meter using encrypted pulse signals. The self-test function built into the taxi meter is used to determine in real time whether the frequency of the encrypted pulse signal is abnormal. If so, a signal abnormality warning message is generated. The anomaly determination module 40 is used to collect real-time taxi operation data, input the real-time taxi operation data and the signal anomaly warning information into the trained taxi monitoring and warning model, and obtain the judgment result of the taxi's abnormal operation behavior.

[0034] Figure 3 This is a schematic diagram of the first page of the taxi monitoring and early warning system according to an embodiment of the present invention, as shown below. Figure 3 As shown, this interface serves as the regulatory access point for the transportation management department. It focuses on specific areas within the "XX District" and specific time periods from "2025-04-10 00:00:00 to 2025-04-10 23:59:59," filtering and displaying taxi data suspected of being illegally equipped with "small motors" for cheating. It supports operations such as querying, resetting, and exporting, facilitating regulators to quickly identify high-risk vehicles and providing visual evidence for precise law enforcement.

[0035] The interface presents three suspicious vehicle records in a table format, with the core fields corresponding to the key judgment dimensions of the monitoring model of this invention: Company Name: Identify the operating company to which the taxi involved belongs, and the corresponding basic operational data collected by the system. License plate number: A unique identifier for suspicious vehicles, serving as a core identifier for law enforcement tracing; Abnormal order volume: The number of orders with mileage / billing abnormalities determined by the model (2 orders for 3 vehicles), which is directly related to core monitoring indicators such as "abnormal comparison between meter mileage and GPS mileage" and "abnormal pulse signal frequency". Credit assessment deduction count: Records of points deducted from a vehicle / driver in the past for illegal operation, corresponding to the credit assessment dimension in the digital profile of people and vehicles built by the system; Number of complaints: The number of passenger complaints against this vehicle, combined with complaint ticket data to help determine suspected cheating; Historical case count: The number of times the vehicle has been investigated and dealt with by law enforcement agencies for illegal operation (including cheating with a "small motor") (currently, all 3 vehicles have 0 cases), used to distinguish the regulatory priority of "first suspected case" and "repeated offense".

[0036] The data on this interface comes directly from the monitoring and early warning process of this invention: Data source: The "abnormal order volume", "number of credit assessment deductions", and "number of complaints" in the table are all from historical / real-time taxi operation data (order data, credit assessment data, complaint work order data, etc.) collected by the system. Analysis logic: The inclusion of a vehicle in the "suspicious database" is the result of multi-level analysis by the model - combining multiple dimensions such as abnormal warning of meter pulse signal, abnormal comparison of GPS and meter mileage, and high-risk marking of historical behavior; Application value: The interface intuitively integrates the core judgment conclusions of the model, avoiding the need for regulators to analyze raw data one by one, greatly improving law enforcement efficiency, and perfectly echoing the core advantages of this invention: "improving regulatory efficiency and reducing law enforcement costs".

[0037] Figure 4 This is a schematic diagram of the second page of the taxi monitoring and early warning system according to an embodiment of the present invention, as shown below. Figure 4 As shown, this study focuses on two abnormal operating orders of a taxi and its driver XX, visually demonstrating the mileage error caused by the cheating behavior of the "little motor" (a driver), providing accurate data support for law enforcement by transportation management departments.

[0038] I. Core Interface Positioning and Data Sources The table data directly originates from the monitoring and early warning process of this invention. It is the result of multi-level analysis by the taxi monitoring and early warning model: Real-time operating data (GPS trajectory, meter mileage) is collected through the vehicle terminal, combined with historical data to construct a digital profile of the driver and vehicle. Through a comprehensive judgment by the rule model layer (mileage threshold comparison), the unsupervised model layer (behavioral pattern recognition), and the historical behavior comparison layer (mileage verification on the same route), orders suspected of being fraudulently obtained are filtered out, forming this detailed table. All data comes from the XX City taxi business system, ensuring authenticity and regional adaptability.

[0039] II. Table Field Meanings and Regulatory Relationships The table contains 12 core fields, each corresponding to a key dimension of "Little Motor" cheating detection. Specific explanations are as follows: Basic identification fields: company name, license plate number, driver's name, clearly identify the parties involved, provide a unique identifier for law enforcement tracing, and link to basic information in the digital profile of people and vehicles; Spatiotemporal information fields: boarding location, boarding time (2025-04-10 22:03:00), drop-off location, drop-off time (2025-04-10 22:21:00), recording the complete spatiotemporal trajectory of the order, which can be combined with urban traffic data to verify the rationality of the trip (such as whether there are mileage anomalies during low-speed periods at night). Key fields for billing and mileage: Payment amount (45 yuan, 58 yuan): Reflects the actual loss to passengers caused by cheating, and the amount is positively correlated with the mileage error; Passenger mileage (14.9 km, 19.9 km): The mileage displayed on the taxi meter, i.e., the data tampered with by "Xiao Mada"; Model-judged mileage (12.8233 km, 16.9902 km): The actual mileage calculated by the system through GPS trajectory and OBD vehicle speed data, providing a benchmark for judging cheating; - Error mileage (2.0767 km, 2.9098 km): The difference between the passenger mileage and the model-judged mileage, which is the core indicator for judging "Xiao Mada" cheating - the error mileage of both orders exceeded the actual mileage by 10% (the error rate of the first order was 16.1%, and the error rate of the second order was 17.1%), far exceeding the threshold of "meter mileage > GPS mileage 10% triggers an anomaly" preset by this invention, directly proving the suspicion of cheating.

[0040] By directly comparing "passenger mileage (tampered data)" with "model-judged mileage (real data)," mileage errors are quantified, breaking the dilemma of "inability to verify the authenticity of meter data" in traditional supervision. The table automatically integrates key order information, eliminating the need for manual verification of original data one by one. Transportation management departments can quickly identify abnormal orders and vehicles involved, echoing the advantages of this invention in "reducing manpower input and shortening the time to detect cheating." The table contains complete data such as timestamps, locations, mileage, and amounts, which, together with abnormal meter pulse signal records and GPS trajectory data, form a traceable electronic evidence chain, solving the problem of "difficulty in obtaining evidence" in existing technologies.

[0041] Figure 5 This is a schematic diagram of the third page of the taxi monitoring and early warning system according to an embodiment of the present invention, as shown below. Figure 5 As shown in the table, the data comes from the taxi industry credit assessment and management platform in XX city. It is the official penalty record of the transportation management department for the violations of taxi drivers and is an important part of the "historical taxi operation data" in the technical solution of this invention. This record is incorporated into the "digital profile of people and vehicles" constructed by the system. It is used to mark the historical violation risk of vehicles in the model training (S102) and abnormal behavior comparison (S104) stages, and to provide historical behavioral basis for the comprehensive judgment of the cheating behavior of "small motors" (taxi drivers), so as to make up for the deficiency of existing technology that "relies only on real-time data for judgment and lacks historical dimension reference".

[0042] The table contains six core fields, each corresponding to both regulatory requirements and technical solutions. Details are as follows: The fields for the involved parties are: company name and license plate number, which clearly identify the operating entity and unique identifier of the vehicle involved in the violation. This corresponds to the technical design of "binding the taximeter and the vehicle terminal" in this invention, ensuring that violation records can be accurately traced to the specific vehicle. Violation Time and Evidence Fields: Incident Time (2025-03-22 22:44:00), Case Number ((2025)0000139), recording the precise timestamp and official filing number of the violation, providing a time anchor for the subsequent system to retrieve GPS trajectory and meter pulse signal data during this period, and assisting in reconstructing the violation scenario; The violation content and penalty fields are: assessment item (failure to use the metering device as required, illegal charging), assessment score (-5), which directly indicates the type of violation - this behavior violates both Article 23 of the "Regulations on the Management of Taxi Operation and Service" which requires "the use of the metering device as required" and falls under the category of "abnormal meter data" that this invention focuses on monitoring; the deduction record quantifies the vehicle's credit risk level. In the "historical behavior comparison layer" of S104, this record will be combined to increase the weight of the abnormal judgment of the taxi and improve the accuracy of identifying cheating behavior.

[0043] Supporting the implementation of technical solutions: This table record is a key component of the "digital profile of people and vehicles" in this invention. In the anomaly judgment in S104, the model will refer to this historical violation record and focus on monitoring vehicles with "previous violations of the meter", thus solving the problem of "indiscriminate judgment and omission of high-risk vehicles" in the existing technology. In line with regulatory requirements: The assessment items in the table directly correspond to Article 48 of the "Regulations on the Management of Taxi Operation and Service," which states that "those who fail to use the metering equipment as required or charge fees illegally shall be fined between 200 and 500 yuan." This invention prevents such violations through technical means (encrypted pulse signals and mileage comparison), and the table records serve as evidence connecting the technical monitoring results with the legal penalties, forming a closed-loop supervision of "technical early warning - violation record - legal penalty." Verification of the necessity of the technology: This record proves that "fare meter violations" are a high-frequency problem in actual operation, further highlighting the necessity of the technical design of this invention, such as "using encrypted pulse signals to prevent tampering and having built-in self-testing alarm function", and specifically solving the pain point of "lack of technical means for fare meter supervision" in existing technologies.

[0044] Figure 6 This is a schematic diagram of the fourth page of the taxi monitoring and early warning system according to an embodiment of the present invention, as shown below. Figure 6As shown in the table, the data comes from the official complaint handling channels in XX city, covering the 12328 transportation service supervision hotline and the 12345 government service hotline. This data is a core component of the "historical taxi operation data" in this invention's technical solution. After being collected and preprocessed in stage S101, this data is integrated into the "digital profile of people and vehicles." In the "historical behavior comparison layer" of S104, it is used to mark the service risk level of vehicles, making up for the blind spot in existing technologies that "only rely on billing data and ignore passenger feedback." This data highly aligns with high-frequency complaint types such as "detours and refusal to carry passengers," reflecting common regulatory pain points in the industry.

[0045] The table contains 7 core fields, each serving the entire process of "risk marking - anomaly association - law enforcement tracing," as detailed below: The fields for the parties involved are: company name, license plate number, and driver's name. This clearly identifies the single entity to which the complaint is directed, echoing the technical design of "binding the meter to the vehicle terminal" in this invention. This ensures that complaint records can be accurately traced to specific vehicles and personnel, avoiding ambiguity in liability determination. Complaint supporting documents fields: complaint order number, acceptance time (June-December 2024), record the official filing number and timestamp of the complaint. The four complaints are distributed within half a year, and two of them occurred in October 2024, reflecting the persistence of the vehicle operation problem. This provides a time anchor for the system to retrieve GPS trajectory and meter pulse signal data for the corresponding period, and can verify whether "the content of the complaint is related to 'Xiao Machi' cheating" (such as "other" type complaints may imply unmentioned mileage abnormalities or overcharging). The complaint content and source fields include three categories: "refusal to carry passengers," "detour," and "other." Among them, "detour" is directly related to the abnormal operation behavior that this invention focuses on monitoring, and it is logically related to the "overcharging" caused by "small motor" cheating (detour + mileage tampering will double the increase in passenger costs), which is consistent with the high frequency of "detour" complaints exposed. The complaint source (12328, 12345) reflects the authority of the complaint, ensures the authenticity of the data, and provides an official basis for subsequent law enforcement verification.

[0046] Supporting the implementation of technical solutions: This table record is a key component of the "digital profile of people and vehicles" in this invention. In the anomaly judgment in S104, the model will refer to this historical violation record - XX taxi has 4 complaints in the past 6 months and has high correlation cheating behaviors such as "driving around". The system will automatically increase the anomaly judgment weight of the vehicle. When the vehicle has abnormal meter pulse signal or mileage error exceeding the threshold (such as meter mileage > GPS mileage 10%), the "confirm anomaly" judgment will be triggered more quickly, solving the problem of "indiscriminate judgment and delayed response to high-risk vehicles" in the existing technology. Aligned with industry regulatory needs: The complaint types (refusal to carry passengers, detours) in the table highly overlap with the violations such as "detours and refusal to carry passengers" exposed in XX City and "detours" exposed in XX City, demonstrating that this invention addresses common pain points in the industry. At the same time, by linking complaint data with billing data and GPS data, a closed-loop supervision of "passenger feedback - technical verification - law enforcement investigation" can be formed, echoing the technical goal of "solving difficulties in obtaining evidence and regulatory lag" in the patent disclosure. Verifying the necessity of the technology: The vehicle received multiple complaints within six months, indicating that traditional regulatory methods are insufficient to eradicate its operational problems. However, this invention, through a dual determination of "complaint data + technical monitoring," can accurately pinpoint whether there is any "small motor" cheating. This further highlights the necessity of technical designs such as "encrypted pulse signal anti-tampering and AI model dynamic analysis," providing a reusable solution for taxi supervision in similar cities.

[0047] Figure 7 This is a schematic diagram of the fifth page of the taxi monitoring and early warning system according to an embodiment of the present invention, as shown below. Figure 7 As shown in the table, the data comes from the XX City taxi complaint handling platform, covering official complaint channels such as the 12328 transportation service supervision hotline and the 12345 government service hotline. This data is a core component of the "historical taxi operation data" in the technical solution of this invention. After being collected in stage S101, this data is preprocessed and integrated into the "human and vehicle digital profile." In the "historical behavior comparison layer" of S104, it is used to mark the service risk level of vehicles, thus compensating for the blind spot in existing technologies that "only rely on billing data and ignore passenger feedback."

[0048] The table contains 7 core fields, each corresponding to the "Service Quality - Fraud Risk" association logic, as detailed below: The fields for the subject of the complaint are: company name, license plate number, and driver's name. This clearly identifies the single subject to which the complaint is directed, which corresponds to the technical design of "binding the meter to the vehicle terminal" in this invention, ensuring that the complaint record can be accurately traced back to the specific vehicle and person. Complaint voucher fields: complaint order number, acceptance time (June-December 2024), record the official filing number and timestamp of the complaint, provide a time anchor for the system to retrieve GPS trajectory and meter pulse signal data for the corresponding period, and help verify whether "the content of the complaint is related to 'small motor' cheating" (such as "other" type complaints may imply mileage anomalies that are not explicitly mentioned); The complaint content and source fields include three categories: "refusal to carry passengers", "detour", and "other". Among them, "detour" is directly related to the abnormal operation behavior that this invention focuses on monitoring, and is logically related to "overcharging" caused by "small motor" cheating (detour + mileage tampering will double increase the passenger's cost); the complaint source (12328, 12345) reflects the authority of the complaint and ensures the authenticity of the data.

[0049] Correlation with technical solutions: In the anomaly detection of S104, the model will refer to the records in this table. If a taxi has 4 complaints within 6 months and has engaged in high-correlation cheating behaviors such as "taking detours", the system will automatically increase the anomaly detection weight of the vehicle. When the vehicle has abnormal meter pulse signal or mileage error exceeding the threshold, the "confirm anomaly" judgment will be triggered more quickly, solving the problem of "indiscriminate judgment and delayed response to high-risk vehicles" in the existing technology.

[0050] Figure 7 Focusing on four normal operating orders of taxis and drivers in XX City from April 8th to 10th, 2025, the complete spatiotemporal trajectory and billing information were recorded to provide data support for the construction of the "normal operating behavior baseline". At the same time, it has the "trajectory playback" function to assist law enforcement verification, which is in line with the technical design of "full-process data retention and dynamic comparison of anomalies" of this invention.

[0051] The table data originates from real-time operational data of the XX City taxi business system. After being collected by in-vehicle terminals (GPS devices, meters, OBD modules), it is transmitted to the system in real-time via the "WebService + Node platform" communication technology and SOAP protocol adopted in this invention. In stage S101, it serves as a "normal sample" for model training, and in stage S104, it serves as a "normal operational baseline" for comparison with abnormal orders, addressing the problems of existing technologies such as "lack of standardized normal behavior references and ambiguous abnormal judgment thresholds." The filter bar at the top of the interface (company name, vehicle license plate number, time range, pick-up location) allows supervisors to accurately query the operational records of specific vehicles, improving data retrieval efficiency.

[0052] The table contains nine core fields, each serving the entire process of "normal baseline construction - anomaly comparison - law enforcement verification," as detailed below: Basic identification fields: company name, license plate number, driver's name, clearly identify the operating entity, match with the basic information in the "human and vehicle digital profile" to ensure unique data ownership; Spatiotemporal trajectory fields: boarding location, boarding time (e.g., 2025-04-09 18:32:00), drop-off location, drop-off time (e.g., 2025-04-09 19:05:00), recording the complete spatiotemporal path of the order. Combined with urban traffic data, "reasonable travel time and mileage" can be calculated (e.g., 18:32-19:05 is the evening peak, and the travel time from Gusu District to Wuzhong District is reasonable), providing a normal baseline for the S104 stage "determining whether there is anomalies such as "low speed at night but a surge in mileage"; Billing and Operation Fields: Payment Amount (52-55 RMB) reflects the normal correspondence between "spatiotemporal trajectory and cost" (e.g., the difference in order amount for similar routes and similar time periods is only 3 RMB, which complies with the pricing rules), and can be compared with the deviation of "mileage-amount" for abnormal orders; The "Track Playback" function allows supervisors to retrieve the GPS dynamic trajectory for the corresponding time period to intuitively check whether the vehicle has taken detours, stopped abnormally, etc., solving the pain point of existing technology "relying on static data for evidence collection and lacking dynamic verification".

[0053] The "normal operation data" recorded in this table is used in stage S102 to train the "normal behavior pattern" of the model (such as the unsupervised model layer using this data to establish normal operation files for taxis). In stage S104, if the vehicle experiences a situation where "the same route is recorded with an extra 2 kilometers of mileage and an extra 15 yuan of fare", the model can quickly identify "deviation from the normal baseline" and trigger an abnormal warning, perfectly demonstrating the core advantages of the invention: "dynamic comparison and accurate identification".

[0054] Figure 8 This is a schematic diagram of the physical structure of an electronic device provided in an embodiment of the present invention, such as... Figure 8 As shown, the electronic device 50 includes: a processor 501, a memory 502, and a bus 503; The processor 501 and the memory 502 communicate with each other via the bus 503. The processor 501 is used to call program instructions in the memory 502 to execute the methods provided in the above-described method embodiments, and to execute the methods provided in the embodiments of the present invention.

[0055] This embodiment provides a non-transitory computer-readable storage medium that stores computer instructions, which cause a computer to execute the method provided in this embodiment of the invention.

[0056] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various storage media capable of storing program code, such as ROM, RAM, magnetic disk, or optical disk.

[0057] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can occur depending on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for monitoring and early warning of taxis, characterized in that, include: Collect historical taxi operation data, and construct model training samples based on the historical taxi operation data; The taxi monitoring and early warning model is trained based on the training samples of the model to obtain a trained taxi monitoring and early warning model; The taxi meter uses encrypted pulse signals to transmit mileage fare data. The built-in self-test function of the taxi meter can determine in real time whether the frequency of the encrypted pulse signal is abnormal. If so, a signal abnormality warning message is generated. Real-time taxi operation data is collected, and the real-time taxi operation data and the signal anomaly warning information are input into a trained taxi monitoring and warning model to obtain the judgment result of abnormal taxi operation behavior.

2. The taxi monitoring and early warning method according to claim 1, characterized in that, The historical taxi operation data includes taxi order data, taxi trajectory data, key vehicle database data, credit assessment results data, complaint work order data, urban traffic condition data, and traffic violation data.

3. The taxi monitoring and early warning method according to claim 1, characterized in that, Following the step of collecting historical taxi operation data, the following steps are also included: The historical taxi operation data is preprocessed, including aggregation, cleaning, deduplication, fusion, standardization, and format unification.

4. The taxi monitoring and early warning method according to claim 1, characterized in that, Following the step of collecting historical taxi operation data, the following steps are also included: A digital profile of people and vehicles is constructed based on the historical taxi operation data.

5. The taxi monitoring and early warning method according to claim 3, characterized in that, The process of training the taxi monitoring and early warning model based on the model training samples to obtain a trained taxi monitoring and early warning model includes: The model training samples are divided into a training set and a validation set; The training set is input into the taxi monitoring and early warning model for training; The taxi monitoring and early warning model is used to verify its accuracy and optimize its parameters using the validation set. When the anomaly recognition accuracy of the taxi monitoring and early warning model reaches a preset threshold, the training of the taxi monitoring and early warning model is completed, and a trained taxi monitoring and early warning model is obtained.

6. The taxi monitoring and early warning method according to claim 1, characterized in that, The self-test function built into the taxi meter is used to determine in real time whether the frequency of the encrypted pulse signal is abnormal. If so, a signal abnormality warning message is generated, including: The frequency is compared with a preset reasonable frequency range. If any of the following situations are detected: frequency exceeding the reasonable range, frequency mutation caused by external device forging encrypted pulse signals, or abnormal increase in frequency during traffic congestion, it is determined that the encrypted pulse signal frequency is abnormal, and a signal abnormality warning message including the abnormality type, the time of abnormality occurrence, and the corresponding meter number is generated.

7. The taxi monitoring and early warning method according to claim 4, characterized in that, The taxi monitoring and early warning model includes an input layer, a rule model layer, an unsupervised model layer, a historical behavior comparison layer, a model comprehensive judgment layer, and an output layer. The real-time taxi operation data and the signal anomaly warning information are received through the input layer; By pre-setting thresholds and business logic rules in the rule model layer, hard billing anomalies in taxi operations can be captured; Identify unknown abnormal behaviors that deviate from normal patterns in taxi operations through unsupervised model layers; By comparing the driver's historical billing mileage with the actual mileage on the same road segment based on the digital profile of the driver and vehicle through the historical behavior comparison layer, high-risk operating objects are marked. The model integrates the judgment results of the hard billing anomaly behavior, unknown anomaly behavior and high-risk operation objects through the comprehensive judgment layer, and combines them with the signal anomaly warning information to conduct a comprehensive anomaly analysis; The output layer outputs the results of the judgment on abnormal taxi operation.

8. A monitoring and early warning system for taxis, characterized in that, include: The data acquisition module is used to collect historical taxi operation data and build model training samples based on the historical taxi operation data. The model training module is used to train the taxi monitoring and early warning model based on the model training samples to obtain a trained taxi monitoring and early warning model. The taxi meter detection module is used to transmit the mileage fare data of the taxi meter using encrypted pulse signals. The self-test function built into the taxi meter is used to determine in real time whether the frequency of the encrypted pulse signal is abnormal. If so, a signal abnormality warning message is generated. The anomaly detection module is used to collect real-time taxi operation data, input the real-time taxi operation data and the signal anomaly warning information into the trained taxi monitoring and warning model, and obtain the judgment result of the taxi's abnormal operation behavior.

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 steps of the method as described in any one of claims 1 to 7.

10. A non-transitory computer-readable medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 7.

Images