Special transformer acquisition terminal based on multiple communication channels and automatic centralized meter reading optimization method thereof

Abstract

The invention provides a special transformer acquisition terminal based on multiple communication channels and an automatic meter reading optimization method thereof, and belongs to the technical field of signal transmission, and the terminal comprises a task determination module which is used for determining acquisition task sets at multiple moments based on a self-adaptive intelligent scheduling acquisition algorithm; the path matching module is used for matching a data transmission path according to the real-time evaluation result of the multiple communication channels and the priority of the collection task at the current moment; the interrupt event acquisition module is used for acquiring an acquisition interrupt event based on the data transmission path; and the retransmission module is used for adaptively determining a retransmission strategy to perform corresponding retransmission according to a first historical interruption situation of the acquisition interruption task if the acquisition succeeds. According to the special transformer acquisition terminal based on the multiple communication channels and the automatic centralized meter reading optimization method thereof, a self-adaptive intelligent scheduling acquisition algorithm is introduced, and task setting is more reasonable. The retransmission behavior is dynamically adjusted, the retransmission strategy is more suitable, and the communication in the data acquisition process is more stable and reliable.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of signal transmission, in particular to a special transformer acquisition terminal based on multiple communication channels and an automatic collection and copying optimization method thereof. BACKGROUND

[0002] Under the background of the continuous advancement of smart grid construction, the power user electricity information acquisition system has become an important support platform for realizing accurate metering, load management, line loss analysis and electricity settlement. Among them, the special transformer acquisition terminal (special transformer user electricity information acquisition terminal) as the key acquisition equipment for high-voltage industrial and commercial large users, undertakes the key functions of time freezing of electric energy meter data, real-time calling and measuring, event reporting, remote control, etc., and its acquisition success rate, real-time performance and reliability are directly related to the operation efficiency and service quality of the power marketing automation system.

[0003] The traditional system usually sets the acquisition tasks of a large number of terminals at the same time point (for example, daily 00:00 generates daily freezing data, and 6:00 uploads uniformly), which causes a large number of terminals to access the communication network at the same time, especially in densely populated urban areas or high-load periods of base stations, which easily causes network bandwidth resource shortage, connection timeout, data packet loss and other problems, and even causes "retransmission storm" in severe cases, further aggravating network pressure and reducing the overall acquisition success rate.

[0004] In addition, most of the existing terminals rely on fixed communication channels, and when signal attenuation, base station switching or electromagnetic interference occurs in the area, they cannot switch to the standby channel in time, resulting in data upload failure. Although some terminals support dual-mode communication, the channel switching is mostly manually configured or simple failover, and there is a lack of real-time dynamic evaluation and intelligent matching mechanism for multiple channel states, making it difficult to achieve optimal path allocation according to task priority and channel quality.

[0005] Therefore, there is an urgent need for a special transformer acquisition terminal based on multiple communication channels and an automatic collection and copying optimization method thereof to at least solve the above problems. SUMMARY

[0006] One of the purposes of the present application is to provide a special transformer acquisition terminal based on multiple communication channels and an automatic collection and copying optimization method thereof, which introduces an adaptive intelligent scheduling acquisition algorithm to determine the acquisition task set at multiple time points, thereby improving the rationality of task setting. When executing the acquisition task, the communication link is matched according to the real-time evaluation results of the multiple communication channels and the priority of the acquisition task. After the link is built, data acquisition is performed, and interruption events are monitored in real time. Based on the cooperative selection mechanism of multiple communication channels, the retransmission behavior is dynamically adjusted according to the current environment and historical experience, and the retransmission strategy is more suitable, which further improves the stability and reliability of communication in the data acquisition process.

[0007] The application embodiment provides a special transformer acquisition terminal based on multiple communication channels, which comprises: A task determination module is configured to determine a plurality of acquisition task sets at multiple time points based on an adaptive intelligent scheduling acquisition algorithm. A channel matching module is configured to match a data transmission channel according to a real-time evaluation result of the multiple communication channels and a priority of an acquisition task in the acquisition task set at the current time point. An interruption event acquisition module is configured to acquire an acquisition interruption event based on corresponding data acquisition of the data transmission channel. A retransmission module is configured to, if the acquisition is successful, adaptively determine a retransmission strategy based on a multiple communication channel cooperative selection mechanism and according to a first historical interruption situation of the acquisition interruption task to perform corresponding retransmission.

[0008] The application embodiment provides an automatic batch copy optimization method of the special transformer acquisition terminal based on multiple communication channels, which is applied to the special transformer acquisition terminal based on multiple communication channels and comprises the following steps: Step 1: determining a plurality of acquisition task sets at multiple time points based on an adaptive intelligent scheduling acquisition algorithm. Step 2: matching a data transmission channel according to a real-time evaluation result of the multiple communication channels and a priority of an acquisition task in the acquisition task set at the current time point. Step 3: acquiring an acquisition interruption event based on corresponding data acquisition of the data transmission channel. Step 4: if the acquisition is successful, adaptively determining a retransmission strategy based on a multiple communication channel cooperative selection mechanism and according to a first historical interruption situation of the acquisition interruption task to perform corresponding retransmission.

[0009] Preferably, step 1: determining a plurality of acquisition task sets at multiple time points based on an adaptive intelligent scheduling acquisition algorithm, comprises the following steps: Generating a plurality of terminal groups and corresponding logical priorities according to geographical or logical zoning. Classifying all to-be-executed acquisition tasks by service and setting an allowed execution time window. Determining a start time of a to-be-executed acquisition task according to a logical priority of a terminal group corresponding to an allowed execution time window of a same service type. Summarizing and determining a plurality of acquisition task sets at multiple time points according to a to-be-executed acquisition task corresponding to each start time.

[0010] Preferably, step 2: matching a data transmission channel according to a real-time evaluation result of the multiple communication channels and a priority of an acquisition task in the acquisition task set at the current time point, comprises the following steps: Comprehensively determining a channel score according to an evaluation index and an index weight of the communication channel, wherein the evaluation index comprises signal strength, bit error rate, round-trip delay, historical upload success rate and current traffic cost. According to the preset matching rule, the data transmission path is matched according to the channel score and the priority of the collection task in the current time.

[0011] Preferably, step 3: collecting corresponding data based on the data transmission path, and obtaining a collection interruption event, including: According to the trigger mechanism preset by the standard interruption type, the triggered standard interruption type is determined; The error code, trigger time, used channel and transmitted data volume of the triggered standard interruption type preset are recorded, and the collection interruption event is obtained.

[0012] Preferably, step 4: if the acquisition is successful, based on the multi-communication channel cooperative selection mechanism, the first historical interruption situation of the collection interruption task is determined, and the retransmission strategy is adaptively determined for corresponding retransmission, including: According to the first historical interruption situation of the collection interruption task, the second historical interruption situation within the target time length and the complete collection process containing the third historical interruption situation are determined; According to the complete collection process, a plurality of switching logics and corresponding exponential backoff models are determined; According to the second historical interruption situation, the target switching logic is matched for exponential backoff and incremental retransmission.

[0013] Preferably, according to the complete collection process, a plurality of switching logics and corresponding exponential backoff models are determined, including: According to the complete collection process, a state transition trajectory is determined; a plurality of state transition points are marked on the state transition trajectory, the distance between the state transition points represents the state transition time interval, and the state transition points are marked with interruption time, switching time, channel data before switching, channel data after switching and standby channel data at switching time; The state transition trajectory is clustered to obtain a standard state transition trajectory; The channel selection logic represented by the standard state transition trajectory is taken as the switching logic; According to the backoff time sequence of the state transition points on the standard state transition trajectory, an exponential trend is determined; Based on a random function, according to the introduced scaling factor and the exponential trend, an exponential backoff model is determined.

[0014] Preferably, the state transition trajectory is clustered to obtain a standard state transition trajectory, including: The structured feature vector of the state transition trajectory is extracted, and the structured features at least include interruption times, whether channel switching, switching delay, average backoff time, total time consumption, task priority and geographical partition; According to the structured feature vector, a similarity matrix between trajectories is constructed; According to the similarity matrix between trajectories, state transition trajectories are grouped based on density clustering DBSCAN or hierarchical clustering algorithm to obtain clustering clusters. In each clustering cluster, a state transition trajectory with the minimum distance to the average vector of other cluster members is determined as a standard state transition trajectory.

[0015] Preferably, according to the backoff time sequence of the state transition point on the standard state transition trajectory, an exponential trend is determined, including: According to the backoff time sequence, the adjacent backoff time ratio is calculated. According to the adjacent backoff time ratio, the stability of the trend is counted based on a preset stability determination standard. If the trend is stable, the exponential trend is determined according to the stable ratio. If the trend is unstable, a candidate function library is matched according to the state transition context information; the candidate function library stores preset candidate function templates of different collection scenarios. According to the candidate function template matched according to the state transition context information and the backoff time sequence, function fitting is performed to determine the exponential trend corresponding to the candidate function template with the minimum error.

[0016] Preferably, the scaling factor is dynamically scaled according to the actual success rate of the standard state transition trajectory in the future.

[0017] The present application has the following beneficial effects: The present application introduces an adaptive intelligent scheduling collection algorithm to determine a plurality of collection task sets at different times, thereby improving the rationality of task setting. When performing the collection task, the communication link is matched according to the real-time evaluation results of a plurality of communication channels and the priority of the collection task. After the link is built, data collection is performed, and interruption events are monitored in real time. Based on the multi-communication channel collaborative selection mechanism, the retransmission behavior is dynamically adjusted according to the current environment and historical experience, and the retransmission strategy is more suitable, thereby further improving the stability and reliability of communication in the data collection process.

[0018] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent to those skilled in the art from the description, or can be learned by practice of the present application. The objects and other advantages of the present application can be realized and achieved by the structure particularly pointed out in the application.

[0019] The technical solutions of the present application will be further described in detail below with the help of the accompanying drawings and examples. BRIEF DESCRIPTION OF DRAWINGS

[0020] The accompanying drawings are used to provide a further understanding of the present application, and constitute a part of the specification, together with the embodiments of the present application, to explain the present application, and do not constitute a limitation on the present application. In the drawings: Figure 1A schematic diagram of the special meter acquisition terminal based on multiple communication channels in the embodiment of the present application; Figure 2 A schematic diagram of the automatic set copy optimization method of the special meter acquisition terminal based on multiple communication channels in the embodiment of the present application. DETAILED DESCRIPTION

[0021] The preferred embodiments of the present application are described below in conjunction with the accompanying drawings, and it should be understood that the preferred embodiments described herein are only used to explain and illustrate the present application, and are not used to limit the present application.

[0022] The embodiment of the present application provides a special meter acquisition terminal based on multiple communication channels, as shown in the figure, comprising: Figure 1 A task determination module 1 is configured to determine a collection task set at multiple time points based on an adaptive intelligent scheduling collection algorithm.

[0023] In the embodiment, the adaptive intelligent scheduling collection algorithm is a scheduling method for staggering the execution of large-scale data collection tasks originally concentrated at the same time and in the same region according to time segmentation and geographical (or logical) grouping. This scheduling method can avoid problems such as communication network congestion, packet loss, and retry caused by a large number of terminals reporting at the same time. The collection task set at multiple time points is a list of to-be-executed collection tasks distributed at different time points, and each task contains terminal ID, collection content, and planned execution time, etc. The task determination module 1 specifically performs the following operations: According to the geographical or logical partition, a plurality of terminal groups and their corresponding logical priority are generated.

[0024] In the embodiment, the terminal group is a management unit in which special meter acquisition terminals with the same geographical or logical characteristics are divided. The logical priority is a scheduling weight value assigned to each terminal group, which is used to determine the execution order of the tasks in the time window.

[0025] Business classification is performed on all to-be-executed collection tasks, and an allowed execution time window is set.

[0026] In the embodiment, the business classification is to divide the to-be-executed collection tasks into different business types according to their importance and real-time requirements (such as real-time type, key timing type, and regular periodic type). For example, the business type is alarm type, which needs to be collected in real time, and the set allowed execution time window is within 5 minutes after the alarm time; the business type is daily frozen power, which needs to avoid the morning peak, and the set allowed execution time window is 06:00-07:30; the business type is curve supplement transmission, which can use the idle bandwidth at night, and the set allowed execution time window is 22:00-05:00.

[0027] ​According to the logical priority of the terminal group corresponding to the allowed execution time window of the same service type, the starting moment of the to-be-executed collection task is determined.

[0028] In this embodiment, according to the logical priority of the terminal group corresponding to the allowed execution time window of the same service type, the starting moment of the to-be-executed collection task is determined: for the same type of service task (for example, all daily frozen electricity), the collection order is arranged according to the logical priority of each terminal group within the allowed execution time period (for example, 6:00-7:30).

[0029] According to the to-be-executed collection task corresponding to each starting moment, the collection task set of multiple moments is summarized and determined.

[0030] The path matching module 2 is configured to match the data transmission path according to the real-time evaluation result of the multiple communication channels and the priority of the collection task in the collection task set at the current moment. Specifically, the channel score is determined by comprehensively considering the evaluation indexes and the index weight of the communication channel. The evaluation indexes include signal strength, bit error rate, round-trip delay, historical upload success rate, and current traffic cost. Based on the preset matching rule, the data transmission path is matched according to the channel score and the priority of the collection task in the collection task set at the current moment.

[0031] In this embodiment, the multiple communication channels are multiple uplink communication modes supported by the terminal, which are used to send data to the master station system. In order to determine which channel is most suitable for the current transmission task, the real-time evaluation result can be obtained by dynamically monitoring and scoring the current state of each communication channel. The evaluation indexes include signal strength, bit error rate, round-trip delay, historical upload success rate, and current traffic cost. Each evaluation index corresponds to an index weight, which represents the influence degree of the evaluation index on the final evaluation result of the channel. The higher the index weight, the greater the influence degree. At the same time, the collection tasks are managed in a hierarchical manner according to the importance of the business, and the priority of the collection task is determined. The preset matching rule is: the higher the priority, the higher the response speed and resource allocation priority for the task (for example, preferentially allocating a communication channel with a higher channel score). After the matching is completed, a one-to-one communication link between the communication channel and the collection task is established, that is, the data transmission path is obtained.

[0032] The interrupt event acquisition module 3 is configured to acquire the collection interrupt event based on the data transmission path. Specifically, the standard interrupt type is determined according to the trigger mechanism preset for the standard interrupt type. The error code, trigger time, used channel, and transmitted data volume of the triggered standard interrupt type are recorded to obtain the collection interrupt event.

[0033] In this embodiment, the standard interruption type is a typical communication or collection failure scenario artificially predefined, including: communication timeout, signal loss, verification failure and connection rejection. The preset triggering mechanism is the detection condition and judgment rule set for the standard interruption type, such as: the triggering mechanism of communication timeout is waiting for response greater than 10 seconds after sending request, and the triggering mechanism of signal loss is that 4G module continuously reads RSSI less than -105dBm for 3 times. The preset error code of the standard interruption type is an encoding representing the specific failure reason.

[0034] The retransmission module 4 is used for, if the acquisition is successful, adaptively determining a retransmission strategy based on the multi-communication channel cooperative selection mechanism according to the first historical interruption situation of the collection interruption task, and performing corresponding retransmission.

[0035] In this embodiment, the multi-communication channel cooperative selection mechanism is an intelligent algorithm for automatically selecting the best alternative path from the standby channel when the main channel fails, realizing seamless switching. The collection interruption task is the collection task corresponding to the collection interruption event, and the first historical interruption situation is the failure record of the task or similar task on different channels in history. The adaptive determination of the retransmission strategy: dynamically adjusting the strategy of the retransmission behavior according to the current environment and historical experience.

[0036] The working principle and beneficial effects of the above technical solutions are: The adaptive intelligent scheduling collection algorithm is introduced to determine the collection task set at multiple times, which improves the rationality of task setting. When the collection task is executed, the communication link is matched according to the real-time evaluation result of the multiple communication channels and the priority of the collection task. After the link is built, data collection is performed, and interruption events are monitored in real time. Based on the multi-communication channel cooperative selection mechanism, the retransmission behavior is dynamically adjusted according to the current environment and historical experience, the retransmission strategy is more appropriate, and further, the stability and reliability of communication in the data collection process are improved.

[0037] As shown in Figure 2 The embodiment of the automatic collection optimization method of the special metering collection terminal based on multiple communication channels is provided, which is applied to the special metering collection terminal based on multiple communication channels in the foregoing embodiment, and includes: Step 1: determining a collection task set at multiple times based on an adaptive intelligent scheduling collection algorithm. Step 1 specifically includes: Step 11: generating multiple terminal groups and their corresponding logical priorities according to geographical or logical partitioning; Step 12: classifying all to-be-executed collection tasks by business, and setting an allowed execution time window; Step 13: determining the start time of the to-be-executed collection task according to the logical priority of the terminal group corresponding to the allowed execution time window of the same business type; Step 14: According to the collection task to be executed corresponding to each starting moment, the collection task set of multiple moments is summarized and determined.

[0038] Step 2: According to the real-time evaluation result of the multiple communication channels and the priority of the collection task in the collection task set at the current moment, the data transmission path is matched. Step 2 specifically includes: Step 21: According to the evaluation index of the communication channel and the index weight, the channel score is comprehensively determined, and the evaluation index includes: signal strength, bit error rate, round-trip delay, historical upload success rate and current traffic cost; Step 22: According to the channel score and the priority of the collection task in the collection task set at the current moment, the data transmission path is matched based on the preset matching rule.

[0039] Step 3: Based on the data transmission path, the corresponding data collection is performed to obtain the collection interruption event. Step 3 specifically includes: Step 31: According to the trigger mechanism preset for the standard interruption type, the triggered standard interruption type is determined; Step 32: The error code, trigger time, used channel and transmitted data volume preset for the triggered standard interruption type are recorded, and the collection interruption event is obtained.

[0040] Step 4: If successful, based on the multi-communication channel cooperative selection mechanism, the retransmission strategy is adaptively determined according to the first historical interruption situation of the collection interruption task for corresponding retransmission.

[0041] The working principle and beneficial effects of the above technical solutions are: The adaptive intelligent scheduling collection algorithm is introduced to determine the collection task set of multiple moments, which improves the rationality of task setting. When the collection task is executed, the communication link is matched according to the real-time evaluation result of the multiple communication channels and the priority of the collection task. After the link is built, data collection is performed, and interruption events are monitored in real time. Based on the multi-communication channel cooperative selection mechanism, the retransmission behavior is dynamically adjusted according to the current environment and historical experience, the retransmission strategy is more appropriate, and further, the stability and reliability of communication in the data collection process are improved.

[0042] In one embodiment, step 4: if successful, based on the multi-communication channel cooperative selection mechanism, the retransmission strategy is adaptively determined according to the first historical interruption situation of the collection interruption task for corresponding retransmission, including: Step 41: According to the first historical interruption situation of the collection interruption task, the second historical interruption situation within the target time length and the complete collection process containing the third historical interruption situation are determined.

[0043] In this embodiment, the target time length is the time length between the start time of the interruption task in the collection and the current time. The second historical interruption situation is all interruption situations in the current task of the interruption task in the collection, including each interruption, channel data before each interruption, and channel data after each interruption. The third historical interruption situation is the first historical interruption situation except the second historical interruption situation, and the complete collection process containing the third historical interruption situation is the complete one-time task communication process of the task or the same type of task with interruption in history, including each interruption, channel data before each interruption, and channel data after each interruption.

[0044] Step 42: According to the complete collection process, a plurality of switching logics and corresponding exponential backoff models are determined.

[0045] In this embodiment, the switching logic is a standard state transition trajectory representation channel switching logic obtained by clustering the complete trajectories of initial attempts, interruptions, retries, switching, successes / failures and other state transitions in a plurality of complete collection processes. The exponential backoff model is matched with a differentiated exponential backoff parameter set according to different channel switching behaviors. For example: in the PSM mode of NB-IoT, delay and wake up to retry, the recommended standard exponential backoff is Random(0, )×1s; switching to the highest-scoring backup channel after two consecutive failures, and the recommended fast exponential backoff is Random(0, )×1s.

[0046] Step 43: According to the second historical interruption situation, the target switching logic is matched for exponential backoff and incremental retransmission.

[0047] In this embodiment, the target switching logic is the switching logic corresponding to the standard state transition trajectory with the highest matching similarity determined by matching the trajectories of state transitions such as interruptions, retries, switching, and failures in the second historical interruption situation with the standard state transition trajectories corresponding to the switching logic. According to the exponential backoff model corresponding to the target switching logic and the historical switching times in the second historical interruption situation, exponential backoff is performed. Incremental retransmission refers to retransmitting only the data part that was not successfully transmitted last time, rather than retransmitting all data.

[0048] The working principle and beneficial effects of the above technical solution are as follows: The present application introduces the previous interruption situation (the second historical interruption situation) of the current task of the interruption task in the collection and the complete collection process of the same type of task as the interruption task in the collection. A plurality of switching logics are induced from the complete collection process, and differentiated exponential backoff models are matched according to different channel switching behaviors for exponential backoff and incremental retransmission, which avoids data storm caused by inappropriate backoff, improves the appropriateness of staggered retransmission, and greatly improves the system stability.

[0049] In one embodiment, step 42: determining a plurality of switching logics and their corresponding exponential backoff models according to the complete acquisition process, comprises: Step 421: determining a state transition trajectory according to the complete acquisition process; a plurality of state transition points are marked on the state transition trajectory, the distance between the state transition points represents the state transition time interval, and the state transition points are marked with the interruption time, the switching time, the channel data before switching, the channel data after switching, and the alternative channel data at the switching time.

[0050] In this embodiment, the state transition trajectory is a time sequence path that records all key state changes in the execution process of the complete acquisition process, and the starting point of the state transition trajectory is the starting time of the complete acquisition process, which is constructed according to the historical log data of the complete acquisition process. The channel data is the channel parameter and the channel evaluation information. The alternative channel is the idle channel.

[0051] Step 422: clustering the state transition trajectory to obtain a standard state transition trajectory.

[0052] In this embodiment, a clustering algorithm is used to merge a large number of similar state trajectories into several categories, with the purpose of discovering typical failure response modes and removing noise and abnormal samples. The standard state transition trajectory is a representative trajectory of the response mode, which is used for subsequent matching with the state trajectory corresponding to the second historical interruption situation, so as to determine the appropriate retransmission strategy. Step 422 specifically includes: Step 4221: extracting a structured feature vector of the state transition trajectory, the structured feature at least including the number of interruptions, whether channel switching, switching delay, average backoff time, total time consumption, task priority, and geographical partition.

[0053] In this embodiment, the structured feature vector is a set of numerical features that can be processed by a machine learning algorithm, which is converted from an unstructured state transition trajectory and represented as a vector form, such as: the number of interruptions: 2 times; whether channel switching: 1 for yes and 0 for no; switching delay: 1.5 seconds; average backoff time: 2.3 seconds; total time consumption: 18 seconds; task priority P1=1 (urgent), P2=2 (important), P3=3 (regular), task priority 1; geographical partition, such as A area=1, B area=2, geographical partition 1; the output structured feature vector example: [2, 1, 1.5, 2.3, 18, 1, 1].

[0054] Step 4222: constructing a similarity matrix between trajectories according to the structured feature vector.

[0055] In this embodiment, the inter-trajectory similarity matrix is an N*N matrix, representing the vector distance between any two of the N state transition trajectories, and both the rows and columns of the matrix represent the trajectory number; the matrix element represents the vector distance between the structured feature vector corresponding to the th trajectory and the structured feature vector corresponding to the th trajectory, and the smaller the vector distance, the more similar the trajectories.

[0056] Step 4223: Based on the density clustering algorithm DBSCAN or the hierarchical clustering algorithm, the state transition trajectories are grouped according to the inter-trajectory similarity matrix, and a clustering cluster is obtained.

[0057] Step 4224: In each clustering cluster, the state transition trajectory with the minimum average vector distance from other cluster members is determined as the standard state transition trajectory.

[0058] Step 423: The channel selection logic represented by the standard state transition trajectory is taken as the switching logic.

[0059] In this embodiment, the channel selection logic is extracted from the standard state transition trajectory, and specifically, the rule for the system to decide to replace the communication channel under certain conditions.

[0060] Step 424: According to the backoff time sequence of the state transition points on the standard state transition trajectory, an exponential trend is determined.

[0061] In this embodiment, the backoff time sequence is an ordered sequence of the time waited before each retry, which is extracted from the time interval between the time of interruption and the time of switching marked on the state transition point. The exponential trend is obtained by analyzing the mathematical rule that the backoff time increases exponentially with the number of failures, for example: the ratio of adjacent backoff times is calculated, and if the ratio tends to a certain constant, there is an exponential trend with the base constant. Then, the constant and the number of failures are substituted into the calculation to obtain the specific exponential trend, for example: Specifically, step 424 includes: Step 4241: According to the backoff time sequence, the ratio of adjacent backoff times is calculated. In this embodiment, the ratio of adjacent backoff times is the quotient of the current backoff time and the previous backoff time.

[0062] Step 4242: According to the ratio of adjacent backoff times, the stability of the trend is calculated based on a preset stability determination standard. In this embodiment, the preset stability determination standard is a standard for determining whether the ratio sequence tends to be stable, for example: if the standard deviation is less than 0.8, the trend is stable, otherwise it is not stable.

[0063] Step 4243: If the trend is stable, the exponential trend is determined according to the stable ratio. In this embodiment, if the trend is stable, a simple exponential model is used to determine the exponential trend.

[0064] Step 4244: If the trend is unstable, match the candidate function library according to the state transition context information; the candidate function library stores candidate function templates preset for different collection scenarios; In this embodiment, if the trend is unstable, the candidate function library is matched according to the state transition context information (which describes the environmental parameters and service attributes when each state change occurs, including: the current channel, the RSSI signal strength, whether the channel has been switched, and the task priority). The candidate function templates preset for different collection scenarios are, for example: if the collection scenario is a high-priority task, the retransmission should not wait for too long, and the corresponding preset candidate function template is , is the priority coefficient; if the collection scenario is a poor signal area, more aggressive backoff is required, and the corresponding preset candidate function template is , is the number of interruptions.

[0065] Step 4245: Perform function fitting according to the candidate function template matched according to the state transition context information and the backoff time sequence, and determine the exponential trend corresponding to the candidate function template with the smallest error.

[0066] In this embodiment, the collection scenario is matched according to the state transition context information, and the candidate function template preset for the collection scenario with a matching degree (scenario similarity) greater than or equal to a preset similarity threshold (for example: 50%) is determined as the matched candidate function template.

[0067] Step 425: Determine the exponential backoff model based on the random function according to the introduced scaling factor and the exponential trend.

[0068] In this embodiment, the random function is a Random function, which is used to add random disturbance on the basis of the determined model to prevent multiple terminals from synchronously retrying, for example: Random(0, ). The scaling factor is a coefficient for adjusting the exponential growth rate, which makes the model closer to the actual data. The scaling factor can be set as a fixed value, or can be dynamically scaled according to the actual success rate of the future statistical standard state transition trajectory, for example: Random(0, ) is the finally determined exponential backoff model, is the scaling factor.

[0069] The working principle and beneficial effects of the above technical solution are: The application visualizes the complete acquisition process data to obtain the state transition trajectory, clusters the standard state transition trajectory, and selects a representative standard state transition trajectory in the cluster.

[0070] Then, the channel selection logic represented by the standard state transition trajectory is used as the switching logic, and the exponential trend is further determined according to the backoff time sequence of the state transition points on the standard state transition trajectory. When determining the exponential trend, the transition from a single simple exponential model to scene perception and dynamic adaptation is realized, and different backoff strategies are adopted for different types of acquisition tasks and different network environments, which significantly improves the intelligent level and communication success rate of the retransmission mechanism. Uncertainty is introduced through a random function, so that even if multiple terminals fail at the same time, their retry times are distributed, a scaling factor for autonomous learning is introduced, and the same exponential structure can adapt to different environments. By combining the scaling factor, the exponential trend and the random function, dynamic adaptation of complex acquisition interruption scenarios is realized, and the scene adaptability is significantly improved.

[0071] Obviously, those skilled in the art can make various modifications and variations to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the claims of the present application and their equivalent technologies, the present application also intends to include these modifications and variations.

Claims

1. A terminal for collecting data from a plurality of communication channels, characterized in that, The application relates to a method for adaptive intelligent scheduling collection algorithm, comprising the following steps: a task determination module is used for determining a collection task set at multiple time points based on the adaptive intelligent scheduling collection algorithm; a channel matching module is used for matching a data transmission channel according to real-time evaluation results of multiple communication channels and priorities of collection tasks in the collection task set at the current time point; an interruption event acquisition module is used for acquiring a collection interruption event based on corresponding data collection of the data transmission channel; a retransmission module is used for adaptively determining a retransmission strategy based on a multi-communication channel cooperative selection mechanism according to a first historical interruption situation of the collection interruption task to perform corresponding retransmission if the acquisition is successful.

2. The automatic collection optimization method based on multi-communication channel of the terminal is characterized in that, The application relates to a method for adaptive intelligent scheduling collection algorithm, comprising the following steps: a task determination module is used for determining a collection task set at multiple time points based on the adaptive intelligent scheduling collection algorithm; a channel matching module is used for matching a data transmission channel according to real-time evaluation results of multiple communication channels and priorities of collection tasks in the collection task set at the current time point; an interruption event acquisition module is used for acquiring a collection interruption event based on corresponding data collection of the data transmission channel; a retransmission module is used for adaptively determining a retransmission strategy based on a multi-communication channel cooperative selection mechanism according to a first historical interruption situation of the collection interruption task to perform corresponding retransmission if the acquisition is successful.

3. The method of claim 2, wherein the method further comprises: Step 1: based on the adaptive intelligent scheduling collection algorithm, a collection task set at multiple time points is determined, comprising the following steps: a plurality of terminal groups and corresponding logical priorities are generated according to geographical or logical partitioning; service classification is performed on all to-be-executed collection tasks, and an allowed execution time window is set; the starting time of the to-be-executed collection task is determined according to the logical priority of the terminal group corresponding to the allowed execution time window of the same service type; the collection task set at multiple time points is determined by summarizing the to-be-executed collection task corresponding to each starting time.

4. The method for automatic batch optimization of a multi-communication channel based end-point meter reading terminal according to claim 2, wherein, Step 2: according to the real-time evaluation results of the multiple communication channels and the priorities of the collection tasks in the collection task set at the current time point, the data transmission channel is matched, comprising the following steps: channel scores are comprehensively determined according to the evaluation indexes and index weights of the communication channels, and the evaluation indexes include signal strength, bit error rate, round-trip delay, historical upload success rate and current traffic cost; based on a preset matching rule, the data transmission channel is matched according to the channel scores and the priorities of the collection tasks in the collection task set at the current time point.

5. The method for automatic batch optimization of a multi-communication channel based end-point meter reading terminal according to claim 2, wherein, Step 3: based on the data transmission channel, corresponding data collection is performed to acquire a collection interruption event, comprising the following steps: the triggered standard interruption type is determined according to the trigger mechanism preset for the standard interruption type; error codes, trigger time, used channels and transmitted data volume of the triggered standard interruption type preset are recorded to obtain the collection interruption event.

6. The method for automatic batch optimization of a multi-communication channel based end-point meter reading terminal according to claim 2, wherein, Step 4: if the acquisition is successful, a retransmission strategy is adaptively determined based on the multi-communication channel cooperative selection mechanism according to the first historical interruption situation of the collection interruption task to perform corresponding retransmission, comprising the following steps: the second historical interruption situation within a target time length and a complete collection process containing a third historical interruption situation are determined according to the first historical interruption situation of the collection interruption task; a plurality of switching logics and corresponding exponential backoff models are determined according to the complete collection process; the exponential backoff and incremental retransmission are performed according to the target switching logic matched with the second historical interruption situation.

7. The method for automatic batch optimization of a multi-communication channel based end device collection terminal according to claim 6, wherein, According to the complete acquisition process, a plurality of switching logics and their corresponding exponential backoff models are determined, including: According to the complete acquisition process, a state transition trajectory is determined; a plurality of state transition points are marked on the state transition trajectory, the distance between the state transition points represents the state transition time interval, and the state transition points are marked with interruption time, switching time, channel data before switching, channel data after switching, and alternative channel data at switching time; The state transition trajectory is clustered to obtain a standard state transition trajectory; The channel selection logic represented by the standard state transition trajectory is used as the switching logic; According to the backoff time sequence of the state transition points on the standard state transition trajectory, an exponential trend is determined; Based on a random function, the exponential backoff model is determined according to the introduced scaling factor and the exponential trend.

8. The method for automatic batch optimization of a multi-communication channel based end-point meter reading terminal according to claim 7, wherein, The state transition trajectory is clustered to obtain a standard state transition trajectory, including: Extracting a structured feature vector of the state transition trajectory, the structured features at least including interruption times, whether channel switching, switching delay, average backoff time, total time consumption, task priority, and geographical partition; According to the structured feature vector, a similarity matrix between trajectories is constructed; Based on the density clustering DBSCAN or hierarchical clustering algorithm, the state transition trajectories are grouped according to the similarity matrix between trajectories to obtain clustering clusters; In each clustering cluster, the state transition trajectory with the minimum average vector distance to other cluster members is determined as the standard state transition trajectory.

9. The method of claim 7, wherein the method further comprises: According to the backoff time sequence of the state transition points on the standard state transition trajectory, an exponential trend is determined, including: According to the backoff time sequence, the adjacent backoff time ratio is calculated; Based on a preset stability determination standard, the trend stability is calculated according to the adjacent backoff time ratio; If the trend is stable, the exponential trend is determined according to the stable ratio; If the trend is unstable, the candidate function library is matched according to the state transition context information; the candidate function library stores candidate function templates preset for different acquisition scenarios; According to the candidate function template matched according to the state transition context information and the backoff time sequence, the function fitting is performed to determine the exponential trend corresponding to the candidate function template with the minimum error.

10. The method of claim 7, wherein the method further comprises: The scaling factor is dynamically scaled according to the actual success rate of the future statistical standard state transition trajectory.

Images