Data acquisition interaction method of electric energy meter data terminal

Abstract

The invention discloses a data acquisition interaction method of an electric energy meter data terminal. The data acquisition interaction method comprises the steps of instruction awakening and parameter initialization, dynamic sampling of environment perception, anomaly detection driven by spatial-temporal characteristics, multi-path quantum-classical hybrid encryption transmission and feedback retransmission and cache synchronization. According to the method, the sampling frequency is dynamically adjusted by fusing power load and environmental fluctuation data, a Mixer-Transform architecture is adopted to extract spatial-temporal characteristics, an abnormal threshold is updated in combination with EWMA, quantum or national secret encryption is selected according to a path quality score, fragmented parallel transmission is carried out, and hierarchical cache synchronization is carried out according to data timeliness. According to the method, the sudden change response speed is increased, the abnormal false alarm rate and the transmission packet loss rate are reduced, the data synchronization delay after network recovery is reduced, and the method is suitable for accurate acquisition and secure transmission of electric energy data in a complex environment.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electric energy metering, in particular to a data acquisition and interaction method of an electric energy meter data terminal, which is especially suitable for accurate acquisition, safe transmission and abnormal detection of electric energy data in complex environments. BACKGROUND

[0002] With the rapid development of smart grids, the electric energy meter data terminal as a key node between the power grid and users, the timeliness, accuracy of data acquisition and the safety of transmission directly affect the scheduling efficiency of the power grid and the level of power management.

[0003] In the prior art, the acquisition and transmission method of the electric energy meter data terminal has many defects: in the aspect of data acquisition, most of them only rely on the absolute value of the power load to adjust the sampling strategy, without considering the influence of environmental factors (such as temperature and humidity, vibration, etc.) on the acquisition accuracy, resulting in large deviation of the collected data when the environment fluctuates greatly; the abnormal detection mostly uses a fixed proportion threshold, which is difficult to adapt to the change of data characteristics in different time and space scenes, and has a high false alarm rate; the data transmission mostly uses single-path switching mode, and the encryption means is single, which is easy to cause data loss or leakage when the network quality is poor or attacked; the cache synchronization mechanism is simple, only data storage is performed, without considering the timeliness of data, resulting in high data synchronization delay after network recovery.

[0004] Therefore, there is an urgent need for a method that can combine environmental perception to realize dynamic sampling, use efficient abnormal detection means, have high security transmission and intelligent cache synchronization, to solve the problems existing in the prior art. SUMMARY

[0005] In view of the shortcomings of the prior art, the purpose of the present application is to provide a data acquisition and interaction method of an electric energy meter data terminal to improve the adaptability of data acquisition, reduce the false alarm rate of abnormal data, enhance the security and reliability of data transmission, and improve the cache synchronization efficiency.

[0006] To achieve the above-mentioned purpose, the present application adopts the following technical solutions: A data acquisition and interaction method of an electric energy meter data terminal, comprising the following steps: S1: instruction wake-up and parameter initialization: the electric energy meter data terminal responds to the acquisition instruction from the remote control center to wake up from the low-power sleep state, parses the acquisition period and acquisition type in the instruction, and initializes the acquisition parameters; S2: dynamic sampling based on environmental perception: collecting power load data and environmental fluctuation data through current sensors, voltage sensors and environmental sensors, dynamically adjusting the sampling frequency based on the load change rate prediction model and the environmental coupling model, specifically meeting:

[0007] wherein is the load sensitive coefficient (dimensionless), representing the influence weight of the electric load on the sampling frequency; is the environmental fluctuation sensitive coefficient (dimensionless), representing the influence weight of the environmental fluctuation on the sampling frequency; is the real-time electric load power value (unit: kW); is the environmental data standard deviation (calculated according to temperature and humidity / vibration sensor data); S3: Spatiotemporal feature driven anomaly detection: using Mixer-Transformer architecture to extract spatiotemporal correlation features from collected data, dynamically updating the anomaly judgment threshold through Exponential Weighted Moving Average (EWMA), and eliminating abnormal data deviating in spatiotemporal dimension; S4: Multi-path quantum-classical hybrid encryption transmission: selecting encryption algorithm according to transmission path quality score, enabling quantum key distribution (QKD) encryption for non-fiber transmission and score > 0.8, otherwise enabling national encryption SM9 encryption; and transmitting data fragments through the two paths with the highest score in parallel; S5: Feedback retransmission and cache synchronization: receiving the confirmation signal from the remote control center, if the transmission fails and the retry count is not exceeded, switching the path and increasing the sampling frequency for retransmission; if the retry is exceeded, storing the data in the local cache and marking the spatiotemporal stamp, and synchronously uploading after the network is restored.

[0008] Further, the load change rate prediction model in S2 adopts ARIMA time series algorithm, specifically including: the standard deviation of the load data in the time window and the historical standard deviation to generate an adaptive threshold , wherein α is 0.6-0.8 and β is 0.05-0.3, when the real-time load change rate exceeds , the sampling frequency is increased to 1.5 times the baseline value.

[0009] Further, the Mixer-Transformer architecture in S3 includes: the temporal convolution module (TemporalConv1d) captures the periodic fluctuations of the load, the feature mixing module (MLP) analyzes the correlation of voltage, current and environmental parameters, and outputs the spatiotemporal feature vector , and the anomaly judgment formula is:

[0010] wherein The historical data is dynamically updated by EWMA.

[0011] Further, the multi-path parallel transmission in S4 satisfies: After data fragmentation, the data is transmitted synchronously through the power line carrier path and the wireless communication path (LoRa / NB-IoT) respectively, and the receiving end recombines the complete data packet based on the data window marked by P0 / P1 code.

[0012] Further, the path quality scoring formula in S4 is:

[0013] Wherein:

[0014] Further, the cache synchronization mechanism in S5 includes: The local cache data is sorted by time and space stamp, and after network recovery, the time-sensitive data is preferentially uploaded (time-sensitive data>environmental fluctuation sensitive data>regular data).

[0015] Further, the environmental sensor includes a temperature and humidity sensor and a vibration sensor, and the environmental fluctuation standard deviation The calculation period is synchronized with the collection period.

[0016] Further, the specific way of dynamically updating the abnormal threshold in S3 is:

[0017] Wherein The historical deviation attenuation factor is 0.85-0.95.

[0018] Further, when the wireless communication path is switched, if the packet loss rate is greater than 10%, the power line carrier path is forced to be enabled to transmit the fragmented main data packet.

[0019] The application also provides an electric energy meter data terminal, comprising: An environmental perception module (including temperature and humidity, vibration sensors), a quantum encryption chip (supporting QKD and SM9), and a Mixer-Transformer coprocessor, which are used to execute any of the above methods.

[0020] One or more technical solutions provided in the embodiments of the application have at least the following technical effects or advantages: By dynamically sampling the environment, the sampling frequency is adjusted in combination with the power load data and the environmental fluctuation data, which improves the mutation response speed compared with the existing method which only relies on the absolute value of the load.

[0021] Adopt the space-time characteristic driven anomaly detection, use Mixer-Transformer architecture to extract features and pass through EWMA dynamic update threshold, and the false alarm rate is improved.

[0022] Multi-path quantum-classical hybrid encryption transmission, combined with path quality score to select encryption algorithm and carry out data fragmentation parallel transmission, make packet loss rate reduce.

[0023] Feedback retransmission and buffer synchronization mechanism, hierarchical upload according to data timeliness, data synchronization delay reduces after network recovery. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1 The system overall architecture diagram of the present application is shown in the figure. Figure 2 The overall flow chart of the present application is shown in the figure. DETAILED DESCRIPTION

[0025] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs; the terminology used in the description herein is for describing particular embodiments only and is not intended to be limiting of the application; the use herein of the terms "and / or" includes a set of one or more associated listed items.

[0026] As shown in Figure 1 and Figure 2 , the embodiment provides a data acquisition interaction method of an electric energy meter data terminal, comprising the following steps: S1: instruction wake-up and parameter initialization: the electric energy meter data terminal is in a low-power sleep state normally, and when a remote control center sends an acquisition instruction, the terminal is woken up. The instruction contains an acquisition period (such as 8:00-20:00 every day) and an acquisition type (such as active power and reactive power), and the terminal initializes acquisition parameters after parsing, including a reference sampling frequency (such as 100 Hz), a data transmission protocol and the like.

[0027] S2: dynamic sampling based on environment perception: current data is acquired by a current sensor, voltage data is acquired by a voltage sensor to obtain electric load data , temperature and humidity data are acquired by a temperature and humidity sensor, and vibration data are acquired by a vibration sensor as environment fluctuation data, and a standard deviation of environment data is calculated . The load sensitivity coefficient is 0.03, the environment fluctuation sensitivity coefficient is 0.02, and the sampling frequency is adjusted according to the formula .

[0028] At the same time, the ARIMA time series algorithm is used as the load change rate prediction model, and the standard deviation of the load data in the past 1 hour is taken as the historical same period standard deviation to generate an adaptive threshold When the real-time load change rate exceeds , the sampling frequency is increased to 1.5 times the reference value.

[0029] S3: Spatio-temporal feature driven anomaly detection: The time convolution module (TemporalConv1d) in the Mixer-Transformer architecture adopts 3 layers of convolution layers to capture the periodic fluctuations of the load within 1 hour; the feature mixing module (MLP) contains 2 layers of fully connected layers to analyze the correlation of voltage, current, temperature and humidity, vibration and other parameters, and output a spatio-temporal feature vector .

[0030] The EWMA dynamically updates the anomaly judgment threshold, and the formula is , wherein When , it is determined as abnormal data and is eliminated.

[0031] S4: Multi-path quantum-classical hybrid encryption transmission: Real-time detection of the bandwidth (BW), delay (Delay) and packet loss rate (Loss) of the power line carrier path, LoRa path and NB-IoT path. MaxBW takes 100Mbps, MaxDelay takes 500ms, and the score of each path is calculated according to the formula .

[0032] If the score of a certain path is > 0.8, quantum key distribution (QKD) encryption is enabled; otherwise, national encryption SM9 encryption is enabled. The data is divided into 2 pieces, and transmitted in parallel through the two paths with the highest score (such as the power line carrier path and the LoRa path), and the receiving end recombines the data packet based on the P0 / P1 code marked data window. When the path is optical fiber, QKD encryption is forced to be enabled.

[0033] S5: Feedback retransmission and cache synchronization: After the remote control center receives the data, it returns an acknowledgment signal containing a digital signature. After the terminal verifies the signature, if the transmission is successful, the bandwidth, delay and other parameters of the path are recorded and the path score weight is updated; if the transmission fails and the retry count (such as 2 times) is not exceeded, switch to the second highest scoring path and increase the sampling frequency to 1.2 times the current frequency for retransmission; if the retry is exceeded, store the data in the local 1GB cache and mark the spatio-temporal stamp, and after the network is restored, upload the time-sensitive data (such as real-time power consumption data) first, and then upload the environmental fluctuation sensitive data and regular data.

[0034] When the packet loss rate of the wireless communication path (such as the LoRa path) is greater than 10%, the power line carrier path is forced to transmit the fragmented main data packet.

[0035] The embodiment also provides an electric energy meter data terminal, which comprises an environment sensing module (containing a temperature and humidity sensor and a vibration sensor), a quantum encryption chip (supporting QKD and SM9), and a Mixer-Transformer coprocessor, and can execute the above method.

[0036] The above merely provides the preferred embodiments of the present application and is not intended to limit the present application. The present application can be variously changed and altered by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A data acquisition interaction method of an electric energy meter data terminal, characterized in that, Comprising the following steps: S1: Instruction wake-up and parameter initialization: the electric energy meter data terminal wakes up from the low-power sleep state in response to the collection instruction of the remote control center, analyzes the collection period and collection type in the instruction, and initializes the collection parameters; S2: Dynamic sampling based on environmental perception: collect electric load data and environmental fluctuation data through current sensors, voltage sensors and environmental sensors, dynamically adjust the sampling frequency based on the load change rate prediction model and the environmental coupling model, and specifically meet: wherein is a load sensitivity coefficient, representing the influence weight of the electric load on the sampling frequency; is an environmental fluctuation sensitive coefficient, representing the influence weight of the environmental fluctuation on the sampling frequency; Real-time power value of the power load, unit: kW; For the environmental data standard deviation, calculated from the temperature / humidity / vibration sensor data; S3: Abnormal detection driven by spatiotemporal characteristics: use the Mixer-Transformer architecture to extract spatiotemporal correlation features from the collected data, dynamically update the abnormal judgment threshold value through the exponentially weighted moving average, and eliminate abnormal data deviating from the spatiotemporal dimension; S4: Multi-path quantum-classical hybrid encryption transmission: select the encryption algorithm according to the transmission path quality score, enable quantum key distribution encryption when the transmission is not optical fiber and the score is >0.8, otherwise enable SM9 encryption, and transmit the data fragments through the two paths with the highest score in parallel; S5: Feedback retransmission and cache synchronization: receive the confirmation signal of the remote control center, if the transmission fails and the retry count is not exceeded, switch the path and increase the sampling frequency for retransmission; If the retry is exceeded, store the data in the local cache and mark the spatiotemporal stamp, and synchronize the upload after the network is restored.

2. The method of claim 1, wherein, The load change rate prediction model in S2 uses the ARIMA time series algorithm, specifically including: standard deviation of load data within a time window with historical standard deviation generating adaptive threshold when real-time load change rate exceeds the sampling frequency is raised to 1.5 times the reference value; Wherein: is the historical load standard deviation, calculated based on the load data within the time window; is a historical volatility weight factor, which has a value ranging from 0.6 to 0.8; a reference offset, the reference offset having a value in the range 0.05 to 0.

3.

3. The method of claim 1, wherein, The Mixer-Transformer architecture in S3 includes: The time convolution module captures periodic fluctuations of the load, and the feature mixing module analyzes the correlation of voltage, current and environmental parameters to output a spatiotemporal feature vector The abnormality determination formula is: wherein: spatio-temporal feature vectors, extracted by a Mixer-Transformer architecture; a predicted value of a feature vector; For the dynamic anomaly threshold, the update is by exponentially weighted moving average.

4. The method of claim 1, wherein, The multi-path parallel transmission in S4 meets: After data fragmentation, the data is transmitted synchronously through the power line carrier path and the wireless communication path, and the receiving end recombines the complete data packet based on the P0 / P1 code marked data window.

5. The method of claim 1, wherein, The path quality score formula in S4 is: wherein and QKD encryption is mandatory when the path is fiber. measured bandwidth for the current transmission path; Unit: Mbps; the maximum nominal bandwidth supported by the system; Unit: Mbps; NetworkDelay, unit: ms, for the current transmission path; The allowable maximum delay threshold is 100ms-500ms. Pcurrentrate of packet loss for current transmission path, dimensionless, percentage: 0.0-1.0; Bandwidth weight coefficient, dimensionless, numerical range: 0.2-0.6; Delay weight coefficient is dimensionless, numerical range: 0.1~0.5; Pd is the packet loss rate weight coefficient, dimensionless, numerical range: 0.2~0.

4.

6. The method of claim 1, wherein, The cache synchronization mechanism in S5 includes: The local cache data is sorted by spatiotemporal stamp, and the time-sensitive data is uploaded first after the network is restored, followed by environmental fluctuation sensitive data and regular data.

7. The method of claim 3, wherein, The environmental sensors include temperature and humidity sensors and vibration sensors, and the environmental fluctuation standard deviation... The calculation cycle is synchronized with the data collection period.

8. The method of claim 1, wherein, The specific way of dynamically updating the abnormal threshold value in S3 is: wherein is a history bias decay factor, with a value ranging from 0.85 to 0.95; t is the current time step.

9. The method of claim 4, wherein, When switching the wireless communication path, if the packet loss rate is >10%, the power line carrier path is forced to transmit the fragmented main data packet.

Images