Urban parking lot vacancy data consistency protection and anomaly detection method

Abstract

The embodiment of the application discloses a kind of urban parking lot remaining seat data consistency protection and exception detection method.The application relates to the technical field of intelligent transportation internet of things application service service, and includes: obtaining the real-time remaining seat data of target parking lot, and generating remaining seat update data packet;First detection component and second detection component with equal value, opposite direction and same detection identifier are generated;The first detection component and the second detection component are sequentially inserted into the remaining seat update data packet, and are sent to the target data receiving node via the remaining seat data transmission link;The detection residual value formed after the target data receiving node processes two detection components and processing sequence are obtained;Determine data loss, repeated processing or out-of-order processing exception;Corresponding node's remaining seat update, and execute data resending, deduplication recovery or sequence correction.The application improves the consistency, reliability and fault tolerance of urban parking lot remaining seat data between parking management platform, induction screen, map platform and travel application.

Description

Technical Field

[0001] This application relates to the field of intelligent transportation Internet of Things application service technology, and in particular to a method for protecting the consistency of urban parking lot space availability data and detecting anomalies. Background Technology

[0002] With the development of urban static traffic management and intelligent parking systems, real-time parking space availability data typically needs to be simultaneously pushed to multiple data receiving nodes, including parking management platforms, parking guidance screen controllers, map service platforms, and travel application servers. Vehicle owners, management entities, and urban traffic management platforms all rely on this availability data for parking guidance, parking lot scheduling, and operational monitoring. Therefore, the consistency of availability data across different receiving nodes directly impacts the effectiveness of urban parking guidance and the reliability of parking lot management.

[0003] Existing parking lot availability systems typically generate real-time availability data based on entrance / exit recognition events, parking space occupancy results, or in-lot vehicle statistics, and then push this data to downstream platforms via network interfaces. However, in actual operation, availability update data packets may experience data loss, duplicate processing, or out-of-order processing due to network jitter, interface retries, message queue congestion, cache overwriting, or node processing delays. For example, if a availability update data packet is not received by the parking guidance screen controller, the screen may display old availability slots for an extended period; if a data packet is consumed repeatedly, the availability slots displayed at the target node may shift relative to the source; if two adjacent availability update data packets are processed in reverse order, the target node may experience incorrect jumps within a short period, or even overwrite correct subsequent availability slots.

[0004] Existing methods often rely on querying the current remaining value of the target node or performing manual calibration after discrepancies are found between displayed values ​​on different platforms. These methods typically only detect differences in the current display results, making it difficult to determine whether the discrepancy is caused by data loss, duplicate processing, or out-of-order processing. Furthermore, they cannot prevent abnormal remaining values ​​from continuing to overwrite the publicly displayed value of the target node in the early stages of anomalies. Especially in city-level parking guidance scenarios, multiple receiving nodes may have different data queues and caching mechanisms; relying solely on comparing the final remaining value is insufficient to promptly detect hidden anomalies in the data transmission chain.

[0005] Therefore, there is a need for a method for protecting and detecting the consistency of remaining bit data and anomalies, which can actively detect the integrity and order of the processing of remaining bit update data packets by each target data receiving node without affecting the actual remaining bit display, and promptly suspend updates, resend data, or correct the order when an anomaly is detected. Summary of the Invention

[0006] This application provides a method for protecting the consistency of urban parking space availability data and detecting anomalies, which improves the consistency, reliability and fault tolerance of urban parking space availability data across parking management platforms, guidance screens, map platforms and travel applications.

[0007] This application provides the following solution: According to the first aspect, a method for ensuring consistency and detecting anomalies in urban parking lot vacancy data is provided, comprising: acquiring real-time vacancy data of a target parking lot and generating a vacancy update data packet; generating a first detection component and a second detection component with equal values, opposite directions, and the same detection identifier according to a preset detection cycle; sequentially inserting the first detection component and the second detection component into the vacancy update data packet and sending it to a target data receiving node via a vacancy data transmission link, wherein the target data receiving node includes a parking management platform, a parking guidance screen controller, a map service platform, or a travel application server; acquiring the residual detection value and processing order formed after the target data receiving node processes the two detection components; determining data loss, duplicate processing, or out-of-order processing anomalies based on the residual detection value and processing order; pausing the vacancy update of the corresponding node according to the anomaly type and performing data resending, deduplication recovery, or sequence correction.

[0008] According to one achievable method in an embodiment of this application, generating the remainder update data packet includes: configuring a remainder version number and a previous version number for the real-time remainder data, and encapsulating the remainder version number, the previous version number, and the real-time remainder data into a remainder update data packet, so that the target data receiving node can determine whether the remainder update data packet is continuous based on the previous version number.

[0009] According to one achievable method in the embodiments of this application, both the first detection component and the second detection component carry a detection identifier, a detection sequence number, a component direction marker, and a component effective range marker. The component effective range marker is used to indicate that the corresponding detection component is only written to the detection buffer of the target data receiving node, and not written to the public residual bit display field of the target data receiving node.

[0010] According to one achievable method in an embodiment of this application, the step of sequentially inserting the first probe component and the second probe component into the remainder update data packet includes: inserting the first probe component into the verification field of the current remainder update data packet, inserting the second probe component into the verification field of the next remainder update data packet adjacent to the current remainder update data packet, and ensuring that the two probe components undergo the same interface authentication, message queue, cache forwarding, and node parsing process.

[0011] According to one achievable method in this application embodiment, obtaining the residual detection value and processing order formed after the target data receiving node processes two detection components includes: setting a hidden detection accumulator at the target data receiving node; when the target data receiving node receives the first detection component or the second detection component, writing the corresponding detection component into the hidden detection accumulator and recording the reception time and processing sequence number of the detection component; when both the first detection component and the second detection component with the same detection identifier are processed or the waiting time reaches a preset time threshold, outputting the accumulation result of the hidden detection accumulator as the residual detection value, and determining the processing order according to the reception time and processing sequence number corresponding to the two detection components.

[0012] According to one achievable method in an embodiment of this application, the step of determining data loss, duplicate processing, or out-of-order processing anomalies based on the residual detection value and processing order includes: determining the processing order of the two detection components based on the processing records corresponding to the first and second detection components; determining that the reverse component is lost when the residual detection value is equal to the first detection component and there is no processing record for the second detection component; determining that the forward component is lost when the residual detection value is equal to the second detection component and there is no processing record for the first detection component; determining that duplicate processing exists when any detection component under the same detection identifier has multiple processing records, or when the absolute value of the residual detection value is greater than the absolute value of a single detection component; and determining that out-of-order processing exists when the processing order indicates that the second detection component was processed earlier than the first detection component.

[0013] According to one achievable method in this application embodiment, the preset detection period is dynamically adjusted based on the historical anomaly count, average response delay, and residual position update frequency of the target data receiving node. This includes: counting the number of data loss, duplicate processing, and out-of-order processing anomalies of the target data receiving node within a preset statistical window to determine the historical anomaly count; counting the average time taken by the target data receiving node from receiving the residual position update data packet to returning the probe residual value to determine the average response delay; counting the number of times the target parking lot generates the residual position update data packet within the preset statistical window to determine the residual position update frequency; shortening the preset detection period of the corresponding target data receiving node when the historical anomaly count reaches a preset anomaly count threshold, the average response delay reaches a preset delay threshold, or the residual position update frequency reaches a preset frequency threshold; and extending the preset detection period of the corresponding target data receiving node when the target data receiving node continuously returns zero residual values ​​within a preset time, the processing order is continuous, and the historical anomaly count, average response delay, and residual position update frequency are all lower than the corresponding thresholds.

[0014] According to one achievable method in an embodiment of this application, pausing the remainder update of the corresponding node based on the anomaly type includes: when it is determined that the target data receiving node has a data loss anomaly, marking the target data receiving node as a missing isolation node, stopping the distribution of remainder update data packets after the missing remainder version to the target data receiving node, and writing the real-time remainder data after the missing remainder version into the corresponding pending-distribution cache; when it is determined that the target data receiving node has a duplicate processing anomaly, marking the target data receiving node as a duplicate isolation node, stopping the distribution of remainder update data packets to the target data receiving node, and keeping the current publicly available remainder data of the target data receiving node as the remainder data corresponding to the most recently confirmed trusted remainder version; when it is determined that the target data receiving node has an out-of-order processing anomaly, marking the target data receiving node as an out-of-order isolation node, stopping the direct overwriting of the current publicly available remainder data of the target data receiving node with newly received remainder update data packets, and writing subsequent remainder update data packets into the corresponding rearrangement cache.

[0015] According to one achievable method in this application embodiment, data retransmission is performed, including: when the target data receiving node is marked as a missing isolation node, determining the missing residual version bound to the unprocessed probe component based on the anomaly detection identifier; obtaining the most recently confirmed trusted residual version of the target data receiving node; determining the residual update data packets between the trusted residual version and the current latest residual version as the data packets to be retransmitted; retransmitting the data packets to be retransmitted to the missing isolation node in a continuous order formed by the residual version number and the previous version number; during the retransmission process, continuing to write the newly generated real-time residual data into the corresponding pending-release cache; and restoring the residual update of the target data receiving node after both the data packets to be retransmitted and the residual update data packets in the pending-release cache have been confirmed and processed by the missing isolation node in a continuous version chain.

[0016] According to one achievable method in the embodiments of this application, the order correction is performed as follows: when the target data receiving node is marked as an out-of-order isolation node, the sequence of remaining bit version numbers processed by the target data receiving node is obtained, and the remaining bit update data packets with a processing order earlier than the previous version number are written into the reordering buffer; the remaining bit update data packets in the reordering buffer are reordered according to the remaining bit version number and the previous version number, and after the sorting result forms a continuous version chain, the target data receiving node is controlled to update the remaining bit data again according to the continuous version chain.

[0017] According to the specific embodiments provided in this application, the following technical effects are disclosed: This application inserts a first and a second probe component with equal values, opposite directions, and identical probe identifiers into the remaining space update data packet. This ensures that the probe data and the actual remaining space data reach each target data receiving node via the same remaining space data transmission link. This proactively detects the data reception and processing status of the nodes without altering their publicly displayed remaining space results. By acquiring the residual probe values ​​and processing order, different anomaly types such as data loss, duplicate processing, and out-of-order processing can be distinguished, avoiding the problem of failing to locate the cause of anomalies solely based on final remaining space value comparison. Furthermore, by pausing the remaining space update of the corresponding node upon detecting an anomaly and performing data retransmission or sequence correction, abnormal remaining space data can be prevented from further overwriting the target node's display results. This improves the consistency, reliability, and fault tolerance of urban parking lot remaining space data across parking management platforms, guidance screens, map platforms, and travel applications.

[0018] Of course, any product implementing this application does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A flowchart illustrating the method for protecting the consistency of available parking space data and detecting anomalies in an embodiment of this application; Figure 2 This is a schematic diagram of the redundant bit data transmission system structure provided in an embodiment of this application; Figure 3 This is a schematic diagram of the structure of the remaining bit update data packet provided in an embodiment of this application. Detailed Implementation

[0021] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0022] Figure 1 This document presents a flowchart illustrating a method for protecting the consistency of available parking space data and detecting anomalies in urban parking lots, as provided in an embodiment of this application. Figure 1 As shown, the method may include the following steps: Step 101: Obtain the real-time availability data of the target parking lot and generate an availability update data package.

[0023] Step 102: Generate a first detection component and a second detection component with equal values, opposite directions, and the same detection identifier according to a preset detection cycle.

[0024] Step 103: Insert the first detection component and the second detection component into the remaining bit update data packet in sequence, and send them to the target data receiving node via the remaining bit data transmission link. The target data receiving node includes a parking management platform, a parking guidance screen controller, a map service platform, or a travel application server.

[0025] Step 104: Obtain the residual detection value and processing order formed after the target data receiving node processes the two detection components.

[0026] Step 105: Determine whether the data is lost, duplicated, or out-of-order based on the detected residual value and processing order.

[0027] Step 106: Based on the anomaly type, pause the remaining bit update of the corresponding node and perform data resending, deduplication recovery, or sequence correction.

[0028] As can be seen from the above process, this application inserts a first probe component and a second probe component with equal values, opposite directions, and the same probe identifier into the remaining space update data packet. This ensures that the probe data and the actual remaining space data reach each target data receiving node via the same remaining space data transmission link. Without changing the publicly displayed remaining space results of the nodes, the application actively detects the data reception and processing status of the nodes. By obtaining the residual probe values ​​and processing order, different anomaly types such as data loss, duplicate processing, and out-of-order processing can be distinguished, avoiding the problem of being unable to locate the cause of anomalies by relying solely on the final remaining space value comparison. Furthermore, by pausing the remaining space update of the corresponding node after an anomaly is detected and performing data retransmission or sequence correction, the application prevents abnormal remaining space data from continuing to cover the target node's display results, improving the consistency, reliability, and fault tolerance of urban parking lot remaining space data across parking management platforms, guidance screens, map platforms, and travel applications.

[0029] The following describes in detail each step of the above process and the effects that can be further produced, with reference to the embodiments.

[0030] Step 101 specifically involves: obtaining the real-time availability data of the target parking lot and generating an availability update data packet.

[0031] The real-time availability data of the target parking lot can be obtained from the vehicle recognition results at the entrance and exit, the availability perception results in the parking lot, the records of the parking management system, or the availability algorithm model. It only represents the number of available parking spaces at a certain moment. Figure 2 This is a schematic diagram of the redundant bit data transmission system structure provided in an embodiment of this application.

[0032] In order to ensure that the real-time residual data can be reliably transmitted between multiple target data receiving nodes, the system does not send individual residual values ​​directly. Instead, it encapsulates the real-time residual data into residual update data packets, making it a data object with version relationships and continuity constraints. Figure 3 This is a schematic diagram of the structure of the remaining parking space update data packet provided in the embodiments of this application, showing the setting positions of real-time remaining parking space data, remaining parking space version number, previous version number, target parking lot identifier and verification field in the remaining parking space update data packet.

[0033] When generating a parking space update data packet, the system configures a parking space version number for each real-time parking space data. This version number uniquely identifies the position of the update within the parking space change sequence of the target parking lot. For example, if the version number of the previous parking space update data packet was V100, the update data packet generated after the current parking space change can be configured as V101. Upon receiving this update data packet, the target data receiving node can determine whether the data is new based on the version number, avoiding duplicate writing of old version data or overwriting already updated parking space data.

[0034] Simultaneously, the system also configures a previous version number for each remainder update data packet. This previous version number points to the previous remainder version that the current remainder update data packet should inherit, establishing a sequential relationship between adjacent remainder update data packets. For example, if the current remainder update data packet has a remainder version number of V101 and its previous version number is V100, it means that this remainder update data packet should be processed after V100. If the target data receiving node currently stores the latest version number as V99, and the previous version number of the received remainder update data packet is V100, it indicates that the target data receiving node has not yet processed version V100. The currently received version V101 cannot directly overwrite the publicly available remainder data; it should first request a resend of version V100 or enter a pending state.

[0035] By encapsulating the remainder version number, the previous version number, and the real-time remainder data into a remainder update data packet, the target data receiving node can not only obtain the current remainder value when receiving data, but also determine the position of that remainder value in the overall update sequence. If the previous version number in the data packet received by the target data receiving node matches the locally saved latest version number, it can be determined that the remainder update data packet is continuous with the local data, allowing the remainder to be updated according to the normal process. If the previous version number does not match the locally saved latest version number, it indicates that the remainder update data packet may have been lost, delayed, or arrived out of order. The target data receiving node can refuse to directly overwrite the currently public remainder data and report the version discontinuity status to the remainder platform.

[0036] This processing method solves the problem of traditional remainder push notifications that only send remainder values ​​without determining data continuity. For parking guidance screens, map service platforms, and travel application servers, even if the received remainder values ​​seem reasonable, they may be inconsistent with the source remainder status due to lost intermediate versions. By introducing remainder version numbers and previous version numbers, the target data receiving node can perform continuity verification before writing, thereby identifying data gaps in advance, avoiding skipping intermediate remainder states and directly displaying subsequent remainders, and reducing the risk of remainder errors spreading across multiple platforms.

[0037] Furthermore, this technical feature provides a foundation for subsequent data retransmission and order correction. When the system detects data loss or out-of-order processing anomalies at a target data receiving node, it can determine the range of missing data packets based on the latest confirmed version number of the remaining bit at that node and the previous version number in the abnormal data packets. Subsequently, the system can retransmit the missing data packets in consecutive order of the remaining bit version numbers, or place out-of-order data packets into a reordering buffer for reordering. Thus, the remaining bit update data packets are no longer isolated data messages, but rather continuous state records with a chain relationship, which helps improve the consistency and recoverability of urban parking lot remaining bit data during multi-node transmission.

[0038] Step 102 specifically involves generating a first detection component and a second detection component that have equal values, opposite directions, and the same detection identifier according to a preset detection cycle.

[0039] The core objective of this step is to add a set of identifiable, cancelable, and traceable detection data to the parking space data transmission link without affecting the actual parking space availability display. The system generates a first detection component and a second detection component according to a preset detection cycle. These two components have equal values, opposite directions, and the same detection identifier. Since the two detection components can cancel each other out when both are correctly processed by the target data receiving node, the theoretical processing result should be zero residue. Thus, the system can use the residual detection value to determine whether the target data receiving node has completely processed the same set of detection components without changing the real-time parking space availability data displayed by the target data receiving node.

[0040] The first and second probe components can be understood as a set of link detection signals that travel alongside the actual remaining bit data. Unlike sending heartbeat packets or test packets separately, the probe components in this application are inserted into the remaining bit update data packet and processed through the same interface authentication, message queue, buffer forwarding, and node resolution process as the actual remaining bit data. Therefore, the probe result does not reflect whether the target data receiving node is online, but rather whether the node's reception, parsing, and processing of the actual remaining bit update data packet is complete, repetitive, or out of order. In this way, anomalies can be detected more closely to the actual business link.

[0041] The detection identifier is used to associate the first and second detection components within the same detection period. Since parking lot availability update data packets may be sent continuously, and multiple detection periods may simultaneously be in a waiting-for-feedback state, numerical values ​​alone cannot distinguish different detection groups. Therefore, each group of first and second detection components is configured with the same detection identifier. When processing the detection components, the target data receiving node can write them into the same detection buffer or the same hidden detection accumulator based on the detection identifier. This avoids confusion between detection components from different detection periods, ensuring that the residual detection value accurately corresponds to a specific link detection process.

[0042] The probe sequence number indicates the order in which probe components are processed within the same probe group. Normally, the first probe component should be processed by the target data receiving node before the second. After recording the probe sequence number, the target data receiving node can determine whether the two probe components were processed in the expected order. Even if both the first and second probe components are eventually processed and the residual probe value is zero, there may still be out-of-order situations where the second probe component arrives first and the first probe component arrives later. The probe sequence number can identify processing order anomalies that cannot be detected by residual values ​​alone, thus preventing older residual data from overwriting newer residual data.

[0043] Component direction markers are used to indicate whether a probe component is forward or reverse. For example, the first probe component can be marked as forward, and the second probe component can be marked as reverse. The target data receiving node determines the direction of action of the probe component in the hidden probe accumulator based on the component direction markers. When both probe components in opposite directions are processed correctly, the accumulation result should be restored to zero. If only the forward probe component is processed, the probe residue value will show a forward residue; if only the reverse probe component is processed, the probe residue value will show a reverse residue. Thus, the system can not only know that data loss has occurred, but also determine whether the lost data packet carries the first probe component or the data packet carries the second probe component.

[0044] The component effective range marker is used to limit the writing location and scope of the probe component, and is an important measure in this technical feature to avoid affecting the actual residual position display. When the target data receiving node parses the residual position update data packet, if it recognizes the component effective range marker carried by the probe component, it will only write the probe component to the probe buffer, hide the probe accumulator, or the probe status field, and will not write it to the public residual position display field. In other words, the probe component participates in link detection and anomaly judgment, but does not participate in the calculation of the residual position value seen by vehicle owners, administrators, or third-party platforms.

[0045] By setting a component effective range flag, it is possible to prevent probe components from being mistaken for actual remainder changes. For example, if the first probe component is a positive value and the second probe component is a negative value, and the target data receiving node directly writes them into the public remainder display field, it may cause a short-term increase or decrease in the remainder display. This application limits the probe components to the probe buffer area, ensuring that they only take effect in the hidden detection channel, thereby guaranteeing that the active probing process does not disturb the actual service display.

[0046] The preset detection period is dynamically adjusted based on the historical anomaly count, average response delay, and residual position update frequency of the target data receiving node. This includes: counting the number of data loss, duplicate processing, and out-of-order processing anomalies of the target data receiving node within a preset statistical window to determine the historical anomaly count; counting the average time taken by the target data receiving node from receiving the residual position update data packet to returning the probe residual value to determine the average response delay; counting the number of times the target parking lot generates the residual position update data packet within the preset statistical window to determine the residual position update frequency; shortening the preset detection period of the corresponding target data receiving node when the historical anomaly count reaches a preset anomaly count threshold, the average response delay reaches a preset delay threshold, or the residual position update frequency reaches a preset frequency threshold; and extending the preset detection period of the corresponding target data receiving node when the target data receiving node continuously returns zero residual values ​​within a preset time, the processing order is continuous, and the historical anomaly count, average response delay, and residual position update frequency are all lower than the corresponding thresholds.

[0047] Because different target data receiving nodes have different network environments, interface processing capabilities, caching mechanisms, and service loads, the probability of data loss, duplicate processing, or out-of-order processing anomalies also varies. If a fixed probing period is always used for all target data receiving nodes, unnecessary data processing overhead may occur on stable nodes, while the probing frequency may be insufficient on high-risk nodes, leading to delayed anomaly detection. Therefore, this application adjusts the preset probing period by combining historical anomaly counts, average response latency, and remaining bit update frequency, so that the probing frequency can change with node status and service load.

[0048] The historical anomaly count represents the cumulative anomalies of the target data receiving node during recent operation. The system counts the number of data loss, duplicate processing, and out-of-order processing anomalies occurring at the target data receiving node within a preset statistical window, and uses the statistical result as the historical anomaly count. For example, if a parking guidance screen controller experiences two data losses and one out-of-order processing event within the last 30 minutes, then the historical anomaly count for this node within the statistical window is 3. A higher historical anomaly count indicates a more unstable data transmission link corresponding to the target data receiving node, requiring an increased detection frequency to more quickly detect subsequent anomalies.

[0049] Average response latency reflects the timeliness of the target data receiving node in processing residual bit update packets and returning probe results. The system calculates the average response latency by tracking the time taken by the target data receiving node from receiving the residual bit update packet to returning the probe residue value, and then calculates the average time taken for multiple residual bit update packets. An increase in average response latency usually indicates that the target data receiving node is experiencing slower interface processing, message queue backlog, buffer forwarding delays, or increased node load. In this state, residual bit update packets are more prone to backlog, out-of-order delivery, or timeouts; therefore, it is necessary to shorten the probe cycle and increase the monitoring density of the node's processing status.

[0050] The remainder update frequency indicates the activity level of remainder data changes in a target parking lot within a preset statistical window. The system counts the number of times the target parking lot generates remainder update data packets within the preset statistical window and determines this number as the remainder update frequency. When vehicles frequently enter and exit the target parking lot, the number of remainder update data packets generated increases, requiring the target data receiving node to continuously process more remainder update data packets. At this time, the risks of message queue congestion, cache overwriting, interface retries, and out-of-order processing all increase. Therefore, even if the target data receiving node itself has a low historical anomaly count, the detection cycle can be appropriately shortened based on a higher remainder update frequency.

[0051] In specific adjustments, when the number of historical anomalies reaches a preset anomaly threshold, the average response latency reaches a preset latency threshold, or the remainder update frequency reaches a preset frequency threshold, the system shortens the preset detection cycle of the corresponding target data receiving node. In other words, as long as any indicator suggests a high risk of anomalies for that node or the current business scenario, the system increases the detection frequency, causing the first and second detection components to be inserted into the remainder update data packets more frequently. This method allows for more timely detection of data loss, duplicate processing, or out-of-order processing issues during periods prone to anomalies.

[0052] Conversely, when the target data receiving node continuously returns zero residual values ​​within a preset time, the processing order is continuous, and the historical anomaly count, average response latency, and parking space update frequency are all lower than the corresponding thresholds, it indicates that the target data receiving node has been processing stably recently, and the current parking lot parking space update pressure is low. In this case, the system can extend the preset detection period of the corresponding target data receiving node, reduce the insertion frequency of detection components, reduce the number of detection cache writes and feedback processing burden of the target data receiving node, while still retaining the periodic detection capability.

[0053] This technical feature also provides a direct basis for subsequent anomaly type determination. If the target data receiving node only processes the first probe component within a preset time, the residual value corresponding to the first probe component will be retained in the probe buffer, which can be used to determine that the second probe component is lost. If only the second probe component is processed, the first probe component can be determined to be lost. If a probe component is repeatedly written to the probe buffer, the residual value will deviate from the normal cancellation result of a single probe component, which can be used to determine duplicate processing. If both probe components are processed but the probe sequence numbers are reversed, out-of-order processing can be determined. Therefore, this application can simultaneously identify multiple anomalies such as data loss, duplicate processing, and out-of-order processing using the same set of probe components.

[0054] Step 103 specifically involves: sequentially inserting the first detection component and the second detection component into the remaining bit update data packet, and sending it to the target data receiving node via the remaining bit data transmission link. The target data receiving node includes a parking management platform, a parking guidance screen controller, a map service platform, or a travel application server.

[0055] The core of this step lies in ensuring that the first and second probe components are not sent as independent test messages, but rather embedded within the actual parking space update data packet, and travel through the actual business link along with the actual parking space data. The purpose of this setup is to ensure that the probe components accurately reflect the processing status of parking space data during its external push process. If only test packets are sent separately, even if they reach the target data receiving node normally, it does not guarantee that the actual parking space update data packet has been correctly processed in processes such as interface authentication, message queues, cache forwarding, and node resolution. This application inserts the probe components into the parking space update data packet, giving it the same transmission path and processing conditions as the actual parking space data; therefore, the detection results are more likely to correspond to anomalies in the actual parking space publishing process.

[0056] In practical implementation, the first probe component is inserted into the checksum field of the current residual bit update data packet, and the second probe component is inserted into the checksum field of the next residual bit update data packet adjacent to the current residual bit update data packet. The current residual bit update data packet refers to the residual bit update data packet generated and sent first within the current probe period; the next adjacent residual bit update data packet refers to the residual bit update data packet generated and sent immediately after the current residual bit update data packet. By arranging two probe components in opposite directions within two adjacent residual bit update data packets, a single set of probe components can be used to simultaneously detect whether two adjacent data packets have been completely received and processed by the target data receiving node.

[0057] This design differs from placing the first and second probe components into the same data packet. If the two probe components are in the same data packet, even if the target data receiving node returns zero remnant, it only indicates that the two probe components within that single data packet were processed simultaneously, making it difficult to detect whether there are lost, skipped, or out-of-order packets between adjacent data packets. However, by placing the first probe component into the current remainder update data packet and the second probe component into the next remainder update data packet, the probe remnant value in the target data receiving node will only return to zero if both adjacent data packets are processed correctly. If either data packet is lost, a non-zero remnant will be formed, thus exposing the transmission discontinuity problem between adjacent remainder update data packets.

[0058] For example, the current remainder update data packet carries the remainder version number V100 and inserts the first probe component into the checksum field; the next remainder update data packet carries the remainder version number V101 and inserts the second probe component into the checksum field. Normally, the target data receiving node should process V100 first, then V101, and write both probe components into the same probe buffer or hide the probe accumulator. Since the two probe components have equal values ​​and opposite directions, if both V100 and V101 are processed correctly, they cancel each other out, and the probe residual value is zero. If the target data receiving node only processes V100 and not V101, the probe buffer will retain the residual value corresponding to the first probe component, indicating that the next remainder update data packet may have been lost or not processed.

[0059] Furthermore, ensuring that both probe components undergo the same interface authentication, message queue, cache forwarding, and node resolution process is to guarantee consistency between the probe process and the actual residual data processing process. The interface authentication step verifies whether the data packet has legitimate access permissions; the message queue step queues and distributes residual update data packets; the cache forwarding step temporarily stores and forwards data between different platforms, services, or terminals; and the node resolution process allows the target data receiving node to parse the real-time residual data, version number, probe identifier, and probe component from the data packet. Any packet loss, duplicate consumption, cache overwriting, or out-of-order forwarding in any of these steps may be reflected in the probe residual value or processing order.

[0060] The target data receiving nodes can be parking management platforms, parking guidance screen controllers, map service platforms, or travel application servers. These nodes perform different functions in the urban parking availability publishing system, but all need to receive and process real-time parking availability data. Parking management platforms are typically used for internal parking lot operation management, parking guidance screen controllers control the display of availability on road or in-lot guidance screens, and map service platforms and travel application servers provide parking guidance information to drivers. Because these nodes may be carried by different systems, different interfaces, or different network links, data loss, duplicate processing, or out-of-order processing at a single node can lead to inconsistencies in the availability displayed on different platforms. This application addresses this by sending availability update data packets with inserted probe components to each target data receiving node, thereby enabling the individual node to check the functionality of its own link.

[0061] In this technical feature, the verification field serves to carry the probe component. Inserting the probe component into the verification field does not alter the real-time residual data field itself. When the target data receiving node parses the residual update data packet, it can extract the probe component from the verification field and write it to the probe buffer according to the probe identifier, while writing the real-time residual data into the residual data field according to the normal business process. This ensures that the probe component travels through the entire link along with the actual residual data, while preventing the probe component from directly affecting the publicly displayed residual results.

[0062] By employing the above methods, this application integrates the continuity detection, processing order detection, and node anomaly identification of surplus bit update data packets into the actual surplus bit push process. When both the first and second probe components are processed by the target data receiving node in the expected order, it indicates that adjacent surplus bit update data packets are being processed normally at that node. When the probe residual value is not zero, it can be determined that data loss or duplicate processing exists. When the residual value is zero but the processing order is reversed, it can be determined that out-of-order processing exists. Therefore, the system can identify hidden anomalies in the surplus bit data transmission link in advance without waiting for significant deviations in the surplus bit display on different platforms.

[0063] Step 104 specifically involves: obtaining the residual detection value and processing order formed after the target data receiving node processes the two detection components.

[0064] When processing the remainder update data packet normally, the target data receiving node, in addition to reading the real-time remainder data and updating its own business data, also identifies the first or second probe component carried in the remainder update data packet. Since the probe component should not change the externally displayed remainder value, a hidden probe accumulator is set in the target data receiving node. The probe component is written to this hidden probe accumulator instead of to the publicly displayed remainder field. The hidden probe accumulator can be set in the target data receiving node's memory, cache database, or temporary state table to store the processing results of probe components under the same probe identifier.

[0065] When the target data receiving node receives the first or second probe component, it first reads the probe identifier carried by the probe component and then looks up the corresponding hidden probe accumulator based on the probe identifier. If the corresponding hidden probe accumulator has not yet been established, a new accumulation record is created for the probe identifier. Subsequently, the value of the probe component is written into the hidden probe accumulator for accumulation. For example, if the first probe component is a forward component and the second probe component is a reverse component, and both have the same probe identifier, and the target data receiving node correctly receives and processes both, the accumulation result in the hidden probe accumulator should be zero. If one probe component is not processed, the accumulator will retain the residual value corresponding to the other probe component.

[0066] While writing to the hidden probe accumulator, the target data receiving node also records the reception time and processing sequence number of the probe component. The reception time indicates when the probe component arrives at the target data receiving node, and the processing sequence number indicates the order in which the probe component is parsed, enqueued, or written to the accumulator within the target data receiving node. The reception time primarily reflects the order of arrival of data packets at the layer level, while the processing sequence number primarily reflects the order of business processing within the node. Both can be used individually or in combination to determine whether two probe components are processed by the target data receiving node in the expected order.

[0067] When both the first and second probe components with the same detection identifier are processed, it indicates that the group of probe components has completed a normal closure, and the target data receiving node can output the accumulated result of the hidden probe accumulator as the probe residue value. If the target data receiving node only receives one of the probe components, it cannot wait indefinitely for the other probe component; therefore, a preset time threshold is set. When the waiting time reaches the preset time threshold, even if the other probe component has not yet arrived, the current accumulated result of the hidden probe accumulator is output. This avoids the detection process from being suspended for a long time due to packet loss and can promptly expose component loss anomalies.

[0068] The probe residue value is the final accumulation result of the hidden probe accumulator under the current probe flag. Normally, the first and second probe components have equal values ​​but opposite directions. After both are correctly processed, the probe residue value should be zero. If the probe residue value is not zero, it indicates that at least one probe component was not correctly canceled, which may correspond to data loss, duplicate processing, or misprocessing. For example, if only the first probe component is processed and the second probe component is not, the probe residue value will be the value of the first probe component; if only the second probe component is processed and the first probe component is not, the probe residue value will be the value of the second probe component; if a probe component is written repeatedly, the probe residue value will deviate from the normal cancellation result.

[0069] The processing order is determined based on the reception time and processing sequence number of the two probe components. If the reception time of the first probe component is earlier than that of the second probe component, and the processing sequence number of the first probe component is also earlier than that of the second probe component, then the probe components can be considered to be processed in the normal order. If the processing sequence number of the second probe component is earlier than that of the first probe component, even if the final residual value is zero, it indicates that the target data receiving node may have processed the data out of order. This design can solve the problem that relying solely on residual values ​​cannot detect out-of-order processing, because even if the two components are processed in reverse order, their values ​​may still cancel each other out, but the corresponding remainder update data packets have already undergone abnormal processing order.

[0070] This technical feature can also distinguish between "out-of-order arrival" and "out-of-order processing." In some cases, the reception times of the two probe components are normal, but due to message queue scheduling, buffer overwriting, or multi-threaded processing within the target data receiving node, the later-arriving data packet is written into the business data first. In this case, the reception time may not show an anomaly, but the processing sequence number will. Conversely, if the reception time is reversed, it can be further determined that the anomaly may occur in the network transmission, buffer forwarding, or upstream queue stage. By simultaneously recording the reception time and processing sequence number, the anomaly localization can be made more specific.

[0071] Overall, this technology features the ability to store the cancellation results of probe components by hiding the probe accumulator, record the processing trajectory of probe components by receiving time and processing sequence number, and terminate abnormal waiting processes in a timely manner by setting a preset time threshold. Its technical advantage lies in obtaining residual probe values ​​and processing order that can be used to determine data loss, duplicate processing, and out-of-order processing, without affecting the publicly displayed residual bit information of the target data receiving node. This provides an accurate basis for subsequent pause of residual bit updates, data retransmission, and sequence correction.

[0072] Step 105 specifically involves determining whether data loss, duplicate processing, or out-of-order processing is an anomaly based on the detected residual value and processing order.

[0073] Since the first and second probe components have the same probe identifier, equal values, and opposite directions, when the target data receiving node processes them in the normal order, the two probe components should cancel each other out in the hidden probe accumulator, resulting in a probe residue value of zero. Furthermore, the processing record of the first probe component should precede the processing record of the second probe component. Based on this normal state, it is possible to determine whether there are any abnormalities in the target data receiving node's processing of the residual bit update data packet by checking whether the probe residue value is zero, whether the processing record is complete, and whether the processing order meets expectations.

[0074] In the specific judgment, the system first determines the processing order of the two detection components based on the processing records corresponding to the first and second detection components. The processing record may include a detection identifier, detection sequence number, component direction marker, reception time, processing sequence number, and corresponding remainder version number. Each time the target data receiving node processes a detection component, it generates a processing record. The system can determine the actual processing order of the first and second detection components in the target data receiving node based on the magnitude of the processing sequence number or the combination of reception time and processing sequence number. In this way, even if the two detection components are ultimately canceled out, it can still identify whether there is an error in the processing order.

[0075] When the residual probe value equals the first probe component and there is no processing record for the second probe component, it indicates that the target data receiving node only processed the first probe component and did not process the second probe component with the same probe identifier. Since the second probe component should have canceled out the first probe component, this state indicates that the residual bit update data packet corresponding to the second probe component may not have reached the target data receiving node, or although it arrived, it was not parsed, did not enter the message queue consumption process, or was not written to the hidden probe accumulator. At this time, the system can determine that the reverse component is lost and further presume that there is a risk of loss or missed processing of the residual bit update data packet carrying the second probe component.

[0076] When the probe residual value equals the second probe component, and there is no processing record for the first probe component, it indicates that the target data receiving node only processed the second probe component and not the first probe component. Since the first probe component is typically inserted into the current remainder update data packet, and the second probe component is inserted into the adjacent next remainder update data packet, this state indicates that the current remainder update data packet was not properly processed at the target data receiving node, while the next remainder update data packet has already been processed. This situation may not only correspond to the loss of a forward component but may also indicate that the target data receiving node may skip the preceding remainder version and directly process the subsequent remainder version.

[0077] When multiple processing records exist for any probe component under the same probe identifier, it indicates that the target data receiving node has processed the same probe component multiple times. For example, the same residual bit update data packet may be processed multiple times due to interface retries, repeated consumption of message queues, repeated forwarding of caches, or repeated execution of node parsing. Even if the final probe residual value does not significantly exceed the absolute value of a single probe component, as long as multiple processing records appear for probe components in the same direction under the same probe identifier, it can be considered that there is a duplicate processing anomaly. This setting can avoid missing some duplicate processing cases by relying solely on numerical judgment.

[0078] If the absolute value of the residual probe value is greater than the absolute value of a single probe component, it can also be determined that duplicate processing or multiple erroneous processing has occurred. Normally, the residual value should be zero after two probe components cancel each other out; if only one component is lost, the residual value is usually equal to the value of the single probe component. If the absolute value of the residual value is greater than the absolute value of a single probe component, it indicates that at least one probe component has been written repeatedly, or that additional erroneous probe components have been mixed in under the same probe identifier. Using this judgment rule, the system can identify more serious problems of duplicate consumption, duplicate parsing, or duplicate writing in the target data receiving node.

[0079] The system determines that out-of-order processing exists when the processing order indicates that the second probe component is processed before the first probe component. This judgment does not depend on whether the probe residual value is zero. This is because, in some cases, even if both the first and second probe components are processed, the final sum may still be zero. However, if the second probe component is processed before the first, it indicates that the processing order of adjacent remainder update data packets carrying the two probe components has been reversed at the target data receiving node. For remainder data publication, out-of-order processing may cause a newer remainder state to be displayed first, and then overwritten by an older remainder state, resulting in incorrect remainder display by the target node.

[0080] The advantage of this technology lies in combining residual value detection and processing order for anomaly identification, rather than simply judging whether the final displayed residual values ​​of the target node are consistent. Residual value detection is mainly used to identify missing components and duplicate processing, while processing order is mainly used to identify out-of-order processing. Together, they can respectively address the three types of anomalies: data loss, duplicate processing, and out-of-order processing. This allows the system to clearly identify the cause of the anomaly and provides a basis for subsequent data resending, node isolation, or sequence correction.

[0081] In actual operation, if the target data receiving node returns a zero residual value, and the first probe component is processed before the second probe component, the target data receiving node can be considered to have processed normally within that probe cycle. If a non-zero residual value is returned, the loss type is determined based on the component direction corresponding to the residual value and the missing processing record. If multiple processing records for the same probe component exist, or the residual value is abnormally amplified, it is determined to be duplicate processing. If the residual value is zero but the processing order is reversed, it is determined to be out-of-order processing. Therefore, this application can accurately identify processing anomalies in the parking space update data packet in the target data receiving node without changing the disclosed parking space display field, improving the consistency and recoverability of multi-node publication of parking space data.

[0082] Step 106 specifically involves pausing the remaining bit update of the corresponding node based on the anomaly type, and performing data resending, deduplication recovery, or sequence correction.

[0083] Since the causes of data loss, duplicate processing, and out-of-order processing are different, using the same recovery method for all of them may not accurately restore the residual bit status of the target data receiving node, and may even amplify the residual bit deviation. Therefore, after determining the anomaly type, this application first isolates the abnormal node, and then performs data retransmission, deduplication recovery, or sequence correction according to the anomaly type, so that the abnormal node will no longer receive and publicly display potentially discontinuous or erroneous residual bit data before recovery.

[0084] When a data loss anomaly is detected at the target data receiving node, it indicates that a certain remainder update data packet or its carried probe component has not been properly processed by the target data receiving node. In this case, if subsequent remainder update data packets continue to be sent to this node, the target data receiving node may skip the missing remainder version and directly display the subsequent remainder status, causing a break in the connection between its published remainder data and the source remainder status. Therefore, the system marks this target data receiving node as a missing remainder isolation node and stops sending remainder update data packets after the missing remainder version to it. Simultaneously, the system does not discard newly generated real-time remainder data but writes it to the corresponding pending-release cache, waiting for the missing version to be completed before releasing it sequentially.

[0085] For missing isolation nodes, the system performs data retransmission. Specifically, the system determines the missing residual version bound to the unprocessed probe component based on the anomaly detection identifier. For example, if the first probe component is bound to residual version V100 and the second probe component is bound to residual version V101, and the target data receiving node only returns the residual value corresponding to the first probe component, and there is no processing record for the second probe component, then V101 can be determined as the missing residual version. Subsequently, the system obtains the most recently confirmed trusted residual version of the target data receiving node and determines the residual update data packets from that trusted residual version to the current latest residual version as the data packets to be retransmitted. This setting does not only retransmit a single missing version, but re-forms a continuous version chain from the trusted version, avoiding the omission of residual update data packets that have been generated but not yet confirmed and processed after the missing version.

[0086] During the retransmission process, the system retransmits the data packets to be retransmitted to the missing isolation node in a continuous order formed by the version number of the remaining bits and the previous version number. This continuous order ensures that the target data receiving node processes the earlier version of the remaining bits first, and then processes the newer version, avoiding packet skipping or the old version overwriting the new version during the retransmission process. Meanwhile, while the retransmission is not yet complete, the target parking lot may continue to generate new real-time remaining bit data. Therefore, the system continues to write the newly generated real-time remaining bit data into the pending-release cache, instead of immediately sending it to the missing isolation node. After both the data packets to be retransmitted and the remaining bit update data packets in the pending-release cache have been confirmed and processed by the missing isolation node according to the continuous version chain, the system then resumes the normal remaining bit update for the target data receiving node.

[0087] When a duplicate processing anomaly is identified at a target data receiving node, it indicates that a probe component under the same probe identifier or its corresponding remainder update data packet has been processed multiple times by the target data receiving node. The duplicate processing anomaly differs from the data loss anomaly in that the same incremental remainder update may be applied repeatedly. For example, if the remainder reduction corresponding to the same vehicle entry event is processed twice, it may cause the target node to publish less remainder data than the actual remainder. Therefore, the system marks the target data receiving node as a duplicate isolation node and stops sending incremental remainder update data packets to this node to prevent the duplicate consumption problem from escalating. Simultaneously, the system maintains the target data receiving node's current published remainder data as the remainder data corresponding to the most recently confirmed trusted remainder version to prevent the anomaly from escalating further.

[0088] For duplicate isolated nodes, the system performs deduplication and recovery. First, the system determines the duplicate remainder version number corresponding to the duplicate processing based on multiple processing records under the same probe identifier. Since the probe component is bound to the remainder update data packet, the processing record of a duplicate probe component can reflect that the corresponding remainder version has been processed repeatedly. The system retains the first valid processing record corresponding to the duplicate remainder version number and marks the remaining duplicate processing records as invalid, eliminating the impact of duplicate processing at the data processing record level. This process can be used for subsequent auditing, anomaly tracing, and preventing the target data receiving node from performing state calculations based on duplicate records again.

[0089] After completing the duplicate record identification, the system obtains the latest trusted duplicate version and its corresponding real-time duplicate data from the duplicate data source, and generates a duplicate recovery data packet carrying a state overwrite flag. The state overwrite method is used here instead of simply sending reverse increments because duplicate processing may have affected multiple subsequent calculation processes, and simply offsetting a single duplicate increment may not necessarily restore the correct state. The duplicate recovery data packet carrying the state overwrite flag instructs the target data receiving node to directly overwrite the currently published duplicate data with the real-time duplicate data corresponding to the latest trusted duplicate version from the source, thereby restoring the duplicate-isolated node to a trusted state consistent with the source.

[0090] When an out-of-order processing anomaly is identified at the target data receiving node, it indicates that although all remainder update data packets may be received by the target data receiving node, their processing order is inconsistent with the remainder version chain. For example, a later-generated remainder version may be processed first, followed by a previous remainder version, potentially causing older remainder data to overwrite newer remainder data. Therefore, the system marks this target data receiving node as an out-of-order isolation node, stops directly overwriting the node's current public remainder data with newly received remainder update data packets, and writes subsequent remainder update data packets to the corresponding rearrangement buffer. This method prevents out-of-order data from continuing to directly affect the public remainder display.

[0091] For out-of-order isolation nodes, the system performs order correction. The system obtains the sequence of remaining bit version numbers processed by the target data receiving node and determines whether there are any remaining bit update packets whose processing order is earlier than the previous version number based on the remaining bit version number and the previous version number. For packets with out-of-order risk, the system writes them to the reordering buffer instead of making them effective immediately. Subsequently, the system reorders the remaining bit update packets in the reordering buffer according to the remaining bit version number and the previous version number, restoring them to a continuous version chain. Only after the sorting result can form a complete continuous version relationship will the system control the target data receiving node to update the remaining bit data again according to this continuous version chain.

[0092] The key feature of this technology lies in handling anomaly isolation and recovery separately according to the anomaly type. Data loss anomalies correspond to version gaps, so continuous resending is used; duplicate processing anomalies correspond to duplicate consumption or duplicate writing, so deduplication and trusted state overwriting are used; out-of-order processing anomalies correspond to version order reversal, so cache rearrangement and continuous version chain correction are used. Through this differentiated processing, this application can avoid the problem that a single recovery strategy cannot adapt to different anomaly causes, and improve the accuracy of residual data recovery.

[0093] Overall, this technology, upon detecting an anomaly at a target data receiving node, does not simply stop pushing data or perform manual calibration. Instead, it first marks the node as a missing isolation node, a duplicate isolation node, or an out-of-order isolation node based on the anomaly type, and then performs data retransmission, deduplication recovery, or sequence correction accordingly. Its technical effectiveness lies in preventing erroneous remaining space data from continuing to overwrite publicly displayed fields before the anomaly node is restored, and ensuring the continuity of the remaining space version chain, the identifiability of duplicate records, and the reordering of out-of-order data during the recovery process. This improves the consistency, reliability, and recoverability of urban parking lot remaining space data during the multi-target node publishing process.

[0094] To further illustrate the technical effects of this application, a specific embodiment and its test results are presented below.

[0095] The method described in this application was deployed in the parking space availability data gateway of a commercial complex parking lot. This parking lot has 780 parking spaces. Real-time availability data is generated jointly by entrance / exit recognition cameras and an in-lot availability recognition model, and is simultaneously pushed to the parking management platform, parking guidance screen controller, map service platform, and travel application server. Each time the availability data gateway generates real-time availability data, it configures an availability version number and the previous version number and encapsulates it into an availability update data packet. Every 20 availability update data packets, the system generates a set of probe components. The first probe component is a positive value, and the second probe component is an equivalent reverse value. Both carry the same probe identifier and are inserted into the verification fields of two adjacent availability update data packets. After receiving the availability update data packet, the target data receiving node writes the probe components into a hidden probe accumulator and returns the probe residue value, reception time, and processing sequence number. If a non-zero probe residue value or an abnormal processing order occurs, the corresponding node is marked as a missing isolation node, a duplicate isolation node, or an out-of-order isolation node, and continuous retransmission, deduplication recovery, or sequence correction are performed respectively.

[0096] To verify the effectiveness of this method, a test environment was constructed containing four types of target data receiving nodes. 12,000 consecutive remainder updates were simulated, with 80 instances of data loss, 60 instances of duplicate processing, and 50 instances of out-of-order processing anomalies artificially injected. Test results showed that this method identified 78 data loss anomalies (97.5%), 58 duplicate processing anomalies (96.7%), and 49 out-of-order processing anomalies (98.0%). Regarding anomaly recovery, the remainder consistency rate for missing isolated nodes after continuous version reissue reached 99.6%, for duplicate isolated nodes after trusted state coverage reached 99.4%, and for out-of-order isolated nodes after cache reordering correction reached 99.7%. Throughout the test, probe components were not written to the public remainder display field, and no remainder jumps caused by probe components occurred on the parking guidance screen or map server. This indicates that this application can improve the consistency and anomaly recovery capabilities of multi-node remainder data without interfering with normal remainder publication.

[0097] The technical solutions provided in this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for protecting the consistency of available parking space data and detecting anomalies in urban parking lots, characterized in that, include: Obtain real-time availability data for the target parking lot and generate an availability update data package; Generate a first detection component and a second detection component that are equal in value, opposite in direction, and have the same detection identifier according to a preset detection cycle; The first detection component and the second detection component are sequentially inserted into the remaining bit update data packet and sent to the target data receiving node via the remaining bit data transmission link. The target data receiving node includes a parking management platform, a parking guidance screen controller, a map service platform, or a travel application server. Obtain the residual detection value and processing order formed after the target data receiving node processes the two detection components; Based on the detected residual value and processing order, data loss, duplicate processing, or out-of-order processing anomalies are determined. Based on the anomaly type, pause the remaining bit update of the corresponding node and perform data resending, deduplication recovery, or sequence correction.

2. The method according to claim 1, characterized in that, The generation of the remainder update data packet includes: Configure a remainder version number and a previous version number for the real-time remainder data, and encapsulate the remainder version number, the previous version number and the real-time remainder data into a remainder update data packet, so that the target data receiving node can determine whether the remainder update data packet is continuous based on the previous version number.

3. The method according to claim 1, characterized in that, Both the first and second detection components carry a detection identifier, a detection sequence number, a component direction marker, and a component effective range marker. The component effective range marker is used to indicate that the corresponding detection component is only written to the detection buffer of the target data receiving node, and not written to the public residual bit display field of the target data receiving node.

4. The method according to claim 1, characterized in that, The step of sequentially inserting the first probe component and the second probe component into the remainder update data packet includes: The first probe component is inserted into the verification field of the current remainder update data packet, and the second probe component is inserted into the verification field of the next remainder update data packet adjacent to the current remainder update data packet. Both probe components undergo the same interface authentication, message queue, cache forwarding and node resolution process.

5. The method according to claim 1, characterized in that, The acquisition of the residual detection value and processing order formed after the target data receiving node processes the two detection components includes: Set a hidden probe accumulator at the target data receiving node; When the target data receiving node receives the first or second probe component, it writes the corresponding probe component into the hidden probe accumulator and records the reception time and processing sequence number of the probe component. When both the first and second detection components with the same detection identifier are processed or the waiting time reaches a preset time threshold, the accumulation result of the hidden detection accumulator is output as the detection residual value, and the processing order is determined according to the receiving time and processing sequence number corresponding to the two detection components.

6. The method according to claim 1, characterized in that, The process of determining data loss, duplicate processing, or out-of-order processing anomalies based on the detected residual value and processing order includes: Based on the processing records corresponding to the first and second detection components, determine the processing order of the two detection components. When the residual value of the detection is equal to that of the first detection component and there is no processing record for the second detection component, it is determined that the reverse component is lost. When the residual value of the detection is equal to that of the second detection component and there is no processing record for the first detection component, it is determined that the positive component is lost. When there are multiple processing records for any detection component under the same detection identifier, or when the absolute value of the residual detection value is greater than the absolute value of a single detection component, it is determined that there is duplicate processing. When the processing order indicates that the second detection component is processed before the first detection component, it is determined that there is out-of-order processing.

7. The method according to claim 1, characterized in that, The preset detection period is dynamically adjusted based on the historical anomaly count, average response latency, and residual bit update frequency of the target data receiving node, including: The number of data loss, duplicate processing, and out-of-order processing anomalies of the target data receiving node within a preset statistical window is counted to determine the historical anomaly count. The average response delay is determined by calculating the average time taken from the time the target data receiving node receives the residual bit update data packet to the time it returns the probe residual value. The number of times the target parking lot generates a spare space update data packet within a preset statistical window is counted to determine the spare space update frequency; When the number of historical anomalies reaches a preset anomaly count threshold, the average response delay reaches a preset delay threshold, or the residual bit update frequency reaches a preset frequency threshold, the preset detection period of the corresponding target data receiving node is shortened. When the target data receiving node continuously returns zero residual value within a preset time, the processing order is continuous, and the number of historical anomalies, average response delay, and residual bit update frequency are all lower than the corresponding thresholds, the preset detection period of the corresponding target data receiving node is extended.

8. The method according to claim 2, characterized in that, The step of pausing the remainder update of the corresponding node based on the anomaly type includes: When it is determined that the target data receiving node has a data loss anomaly, the target data receiving node is marked as a missing isolation node, the publication of the missing remaining bit update data packet after the missing remaining bit version to the target data receiving node is stopped, and the real-time remaining bit data after the missing remaining bit version is written into the corresponding pending publication cache area. When it is determined that the target data receiving node has a duplicate processing anomaly, the target data receiving node is marked as a duplicate isolation node, the sending of residual bit update data packets to the target data receiving node is stopped, and the current public residual bit data of the target data receiving node is kept as the residual bit data corresponding to the most recently confirmed trusted residual bit version; When it is determined that the target data receiving node has an out-of-order processing anomaly, the target data receiving node is marked as an out-of-order isolation node, the direct overwriting of the target data receiving node's current public residual bit data with newly received residual bit update data packets is stopped, and subsequent residual bit update data packets are written to the corresponding rearrangement buffer.

9. The method according to claim 8, characterized in that, Data resending is performed, including: When the target data receiving node is marked as a missing isolation node, the missing bit version bound to the unprocessed probe component is determined according to the anomaly detection identifier; Obtain the most recently confirmed trusted residual bit version of the target data receiving node; The remaining bit update data packets between the trusted remaining bit version and the current latest remaining bit version are identified as the data packets to be resent. The missing isolation node is resent the data packet to be resent in the continuous order formed by the remaining version number and the previous version number. During the reissue process, the newly generated real-time remaining data will continue to be written to the corresponding release cache area; Once the missing isolation node has confirmed and processed both the data packet to be resent and the remaining bit update data packet in the buffer to be published according to the continuous version chain, the remaining bit update of the target data receiving node is restored.

10. The method according to claim 8, characterized in that, Execution order correction, including: When the target data receiving node is marked as an out-of-order isolation node, obtain the sequence of remaining bit version numbers that the target data receiving node has processed, and write the remaining bit update data packets whose processing order is earlier than the previous version number into the rearrangement buffer. The remaining bit update data packets in the rearrangement buffer are reordered according to the remaining bit version number and the previous version number. After the sorting results form a continuous version chain, the target data receiving node is controlled to update the remaining bit data according to the continuous version chain.

Images