Heat meter verification data intelligent analysis system based on heat supply multi-parameter simulation

Abstract

The invention relates to the technical field of thermoelectric Internet of Things safety, in particular to a heat meter verification data intelligent analysis system based on heat supply multi-parameter simulation, which comprises a multi-parameter working condition simulation module, a data acquisition and communication module, a verified household heat meter, a data processing and storage module, an intelligent analysis core module and a calibration optimization suggestion module, the multi-parameter working condition simulation module accurately reproduces water quality, temperature, flow velocity and battery electric quantity coupling working conditions; the data acquisition and communication module realizes unified acquisition of heat meter data of multiple brands; the intelligent analysis core module quantifies working condition coupling influence, sorts heat table performance and predicts the service life of the battery through a multivariate regression + SHAP hybrid model, an entropy weight-TOPSIS method and an ARIMA model; the calibration optimization suggestion module generates a dynamic calibration coefficient. According to the invention, all-condition verification, intelligent analysis and accurate optimization of the heat meter are realized, data support is provided for purchase, calibration, operation and maintenance of the heat meter of a heat supply enterprise, and the metering accuracy and the energy utilization efficiency are improved.

Description

Technical Field

[0001] This invention relates to the field of thermoelectric Internet of Things (IoT) safety technology, specifically to an intelligent analysis system for heat meter calibration data based on multi-parameter simulation of heating. Background Technology

[0002] With the rapid development of the centralized heating industry, household heat meters, as core metering instruments for trade settlement, directly affect the vital interests of heating companies and residents in terms of measurement accuracy and operational stability. They also significantly impact the energy conservation, emission reduction, and standardized operation of heating systems. However, the heating industry currently faces numerous prominent problems in the use and calibration of heat meters, severely hindering the industry's high-quality development.

[0003] On the one hand, the actual operating environment of heat meters is complex and variable, and the accuracy of metering is affected by multiple coupled factors, making it difficult for traditional verification methods to fully cover actual operating conditions. Residential heat meters, in their long-term use, not only face the influence of external conditions such as differences in pipe network water quality (e.g., calcium and magnesium ion content, sediment, rust, and other impurities), fluctuations in supply and return water temperatures, and uneven flow rates in the pipes, but also the effects of internal factors such as battery degradation and disassembly / removal wear, resulting in a persistently high failure rate. During the trade settlement statement rectification campaign, a large number of faulty heat meters were detected, and the metering accuracy of heat meters that had not undergone regular verification lacked effective verification. Traditional heat meter verification is mostly based on a single standard operating condition, only testing metering performance under fixed parameters, and cannot simulate the complex scenarios of multiple coupled parameters in actual operation. This leads to a disconnect between verification results and the actual usage state of the heat meters, making it difficult to reveal the independent influence and interaction of various factors on metering errors, and failing to provide a scientific basis for heat meter error correction.

[0004] On the other hand, the heat meter market is rife with numerous brands and varying quality, lacking a unified and standardized performance evaluation system. Furthermore, data collection and analysis methods are outdated, leading to inefficient operation and maintenance management. According to feedback from heating companies, different brands of heat meters exhibit significant differences in failure rates, maintenance difficulty, and environmental adaptability. However, existing evaluations largely rely on manual experience or simple data comparisons, lacking multi-dimensional and quantitative comprehensive evaluation models. This makes it difficult to objectively reflect the actual performance level of heat meters, causing significant difficulties for heating companies in procurement and selection. In addition, the inconsistent communication protocols of different brands of heat meters result in cumbersome data collection processes, poor compatibility, and an inability to achieve unified collection and standardized processing of batch data. Traditional analysis methods are often limited to simple statistics of single parameters, failing to achieve accurate tracing of metering errors, early warning of battery life, and timely fault diagnosis. This leads to a series of problems such as frequent metering disputes, high operation and maintenance costs, and energy waste.

[0005] With the increasing national requirements for energy conservation and emission reduction in the heating industry, and the urgent need for heating companies to improve operational efficiency and service quality, the limitations of traditional heat meter calibration methods in terms of operating condition coverage, data processing, analysis and evaluation are becoming increasingly apparent. There is an urgent need for a heat meter calibration system that can simulate complex multi-parameter coupled operating conditions, realize intelligent data analysis, and provide comprehensive performance evaluation and accurate optimization suggestions to fill the existing technological gap and promote the standardized, scientific and intelligent development of metering management in the heating industry. Summary of the Invention

[0006] The purpose of this invention is to provide an intelligent analysis system for heat meter calibration data based on multi-parameter simulation of heating, so as to solve the increasingly prominent limitations of the heat meter calibration methods mentioned in the background art in terms of operating condition coverage, data processing, analysis and evaluation.

[0007] To achieve the above objectives, the present invention provides the following technical solution: A smart analysis system for heat meter calibration data based on multi-parameter heating simulation includes: The multi-parameter operating condition simulation module is used to dynamically adjust and generate a multi-parameter coupled operating condition environment, including water quality parameters, temperature parameters, flow rate parameters and battery power, in the heating operating condition simulation laboratory according to the preset test plan. The data acquisition and communication module is connected to the multi-parameter operating condition simulation module and the household heat meter under test, and is used to collect operating condition parameter data and metering response data of the household heat meter under test in real time under the multi-parameter coupled operating condition environment. The data processing and storage module is used to receive and store raw data from the data acquisition and communication module, and to clean, align and standardize the data to form a structured test dataset. The intelligent analysis core module, connected to the data processing and storage module, is used to calculate the performance indicators of the tested household heat meters based on the structured test dataset, through built-in analysis models and algorithms, analyze the multi-parameter coupling influence law, and generate a performance evaluation report. The intelligent analysis core module includes at least an operating condition coupling effect analysis unit and a heat meter performance comprehensive evaluation unit. The operating condition coupling effect analysis unit adopts a hybrid model based on multivariate regression and Shapley additive interpretation (SHAP) to quantify the independent and interactive contributions of each operating condition parameter to the heat meter measurement error. The comprehensive performance evaluation unit for heat meters is based on the entropy weight-TOPSIS method to construct a multi-dimensional comprehensive performance evaluation model for heat meters, and outputs the comprehensive performance ranking and adaptability level of each tested heat meter.

[0008] Preferably, the water quality parameter simulation in the multi-parameter operating condition simulation module is achieved by adding a specific concentration of chemical substances to the circulating water, and the concentration change function for simulating water hardness is: ; in, express Ion concentration at time [time] This is the initial concentration. For the target concentration, This function represents a concentration adjustment coefficient related to the addition rate and system volume, enabling precise control and simulation of water quality changes.

[0009] Preferably, the metering response data collected by the data acquisition and communication module includes instantaneous flow rate, cumulative flow rate, supply and return water temperature, calculated heat, and instantaneous fault code; the collected operating condition parameter data includes pipeline pressure, standard meter flow rate, system temperature, flow rate, water conductivity, and battery voltage.

[0010] Preferably, the core algorithm for quantifying the contribution of the impact in the operating condition coupling impact analysis unit is as follows: First, construct the measurement error. With multiple operating parameters Multivariate nonlinear regression model: ; in, For the intercept term, and For regression coefficients, and For nonlinear basis functions, This is the error term; Then, the trained regression model is used as the prediction model for SHAP analysis to calculate the parameters for each operating condition. SHAP value on each sample The mean of the SHAP value This indicates the average marginal contribution of the parameter to the measurement error.

[0011] As a preferred embodiment, the specific implementation steps of the entropy weight-TOPSIS method in the comprehensive performance evaluation unit of the heat exchanger are as follows: Step S1: Construct the evaluation matrix , line representative The evaluated hot list represents Several performance evaluation indicators, including average error, standard deviation of error, number of times the maximum permissible error exceeds the standard, and battery capacity decay rate under different operating conditions; Step S2: Calculate the objective weights of each evaluation index using the entropy weight method. : S21: Normalize the evaluation matrix to obtain the normalized matrix. For positive indicators For negative indicators ; S22: Calculate the... Entropy value of each indicator : ; in, ,like ,but , , ; S23: Calculate the... Weight of each indicator : ; Step S3: Calculate the comprehensive evaluation results using the TOPSIS method: S31: Construct a weighted normalization matrix , ; S32: Determine the ideal solution and negative ideal solution ,in , ; S33: Calculate the Euclidean distance from each heat exchanger to the positive and negative ideal solutions: ; ; S34: Calculate relative proximity : ; in accordance with The values ​​are used to rank all tested heat exchangers based on their overall performance.

[0012] Preferably, the intelligent analysis core module also includes a battery power degradation prediction unit. This unit predicts the remaining service life (RUL) of the hot meter battery based on the collected battery voltage time series data using a time series analysis algorithm. The prediction model adopts a differential autoregressive moving average model, the general form of which is: ; in, for Battery voltage at any given time, For the shift operator, Let be the difference order. It is a constant. These are the autoregressive coefficients. The moving average coefficient is... It is a white noise sequence. and These are the autoregression order and the moving average order, respectively.

[0013] Preferably, the battery power degradation prediction unit predicts the remaining service life. With a preset security threshold When comparing, When this happens, the system automatically generates a warning message and indicates that the heat meter has a potential battery power risk.

[0014] Preferably, the system further includes a calibration optimization suggestion module, which generates dynamic calibration compensation coefficients based on the output of the operating condition coupling effect analysis unit, targeting the systematic errors of a specific brand or model of heat exchanger under specific adverse operating conditions. The calculation formula is as follows: ; in, relative to standard temperature temperature difference, relative to standard flow rate The difference in flow velocity, This is the water hardness value, expressed as CaCO3, in units of... , This is the error sensitivity coefficient unique to this model of heat exchanger, obtained through fitting the influence analysis model.

[0015] Preferably, the data acquisition and communication module supports communication with multiple heat meters of different brands and with different communication protocols. It has a built-in configurable communication protocol library covering mainstream protocols such as RS485, Modbus RTU, LoRa, and NB-IoT. Through protocol parsing and adaptation, it achieves unified data acquisition and standardized parsing.

[0016] Preferably, the performance evaluation report and calibration optimization suggestions generated by the system are output to the asset management system or meter procurement decision support system of the heating company through a standardized data interface. The data interface supports communication methods such as RESTful API and WebSocket to realize data-driven decision support for heat meter procurement selection, replacement upon expiration, on-site calibration and operation and maintenance.

[0017] Compared with the prior art, the beneficial effects of the present invention are: (1) This invention achieves accurate reproduction of multi-parameter coupled operating conditions of the heating system through a multi-parameter operating condition simulation module, enabling dynamic adjustment and precise control of key parameters such as water quality, temperature, flow rate, and battery charge. Combined with quantitative models such as concentration change function, it can comprehensively simulate various actual scenarios from standard operating conditions to extreme operating conditions, making the verification process closer to the actual operating state of the heat meter. At the same time, relying on the multivariate regression and SHAP hybrid model, it can not only quantify the independent influence of each parameter on the measurement error, but also reveal the interactive coupling effect between parameters, filling the technical gap of the vague law of operating condition influence in traditional verification, and providing more comprehensive and in-depth technical support for the analysis of the accuracy of heat meter measurement.

[0018] (2) This invention constructs a fully intelligent heat meter verification and evaluation system, significantly improving the scientific nature of heat meter performance evaluation and the efficiency of operation and maintenance management. The multi-protocol adaptation capability of the data acquisition and communication module enables unified data acquisition and standardized parsing of heat meters of different brands and types, greatly reducing the operational complexity of multi-brand heat meter verification. The entropy weight-TOPSIS comprehensive evaluation algorithm and ARIMA battery life prediction model integrated in the intelligent analysis core module can output objective and fair heat meter performance ranking and accurate warning of remaining service life, solving the problems of traditional evaluation relying on manual experience and lagging fault diagnosis. The dynamic calibration compensation coefficient generated by the calibration optimization suggestion module can specifically correct systematic errors under specific working conditions. Combined with warning information and operation and maintenance suggestions, it forms full life cycle management support for heat meters from procurement and selection, on-site calibration to replacement upon expiration, effectively reducing metering disputes and operation and maintenance costs caused by heat meter failures.

[0019] (3) This invention has strong industry applicability and promotional value, providing a strong guarantee for energy conservation, consumption reduction, and standardized development in the heating industry. The performance evaluation report and operating condition influence law analysis generated by the system are seamlessly connected with the asset management system and procurement decision-making system of heating enterprises through standardized interfaces, realizing a data-driven decision-making model. This helps enterprises accurately select heat meter products with strong adaptability and high stability, and optimize procurement costs. At the same time, the unified heat meter performance evaluation standard and verification process can standardize the market access and quality assessment of heat meters, forcing enterprises to improve product quality and promote the technological upgrading of the industry. In addition, accurate meter calibration and fault early warning can reduce energy waste caused by heat meter metering deviation, improve the energy utilization efficiency of the heating system, meet the development needs of energy conservation and consumption reduction in the heating industry, and have both economic value and social benefits. Attached Figure Description

[0020] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are explained in detail together with the embodiments of the invention, but do not constitute a limitation thereof.

[0021] Figure 1This is a system composition block diagram of the present invention; Figure 2 This is a performance comparison of different brands of household heat meters under standard operating conditions in Embodiment 1 of the present invention. Brand A performed the best, with a relative similarity score of 0.85. Figure 3 To illustrate the influence of various parameters on the heat meter measurement error under extreme operating conditions in Embodiment 2 of the present invention, water hardness and impurity content are the main influencing factors. Figure 4 The performance distribution of different batches of heat meters in Embodiment 3 of the present invention shows that batch 2 has the best overall performance, while two heat meters in batch 1 have poor performance. The meanings of the structural labels in the diagram are as follows: 100, Multi-parameter operating condition simulation module; 200. Data acquisition and communication module; 300. Household heat meters being inspected; 400. Data processing and storage module; 500. Intelligent Analysis Core Module; 510. Operating Condition Coupling Influence Analysis Unit; 520. Heat Surface Performance Comprehensive Evaluation Unit; 530. Battery Capacity Attenuation Prediction Unit; 600. Calibration optimization suggestion module. Detailed Implementation

[0022] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0023] like Figure 1 As shown, the overall architecture of the intelligent analysis system for heat meter calibration data based on multi-parameter simulation of heating in this invention includes a multi-parameter operating condition simulation module 100, a data acquisition and communication module 200, a household heat meter under calibration 300, a data processing and storage module 400, an intelligent analysis core module 500, and a calibration optimization suggestion module 600. Each module communicates through a standardized interface to form a closed-loop workflow of operating condition simulation-data acquisition-processing analysis-optimization output.

[0024] (I) Multi-parameter operating condition simulation module 100 The multi-parameter operating condition simulation module 100 is deployed in the heating operating condition simulation laboratory. It consists of a water quality simulation unit, a temperature regulation unit, a flow rate control unit, a battery power regulation unit, and a central controller. It can dynamically generate preset multi-parameter coupled operating condition environments.

[0025] Water quality simulation unit: A high-precision metering pump adds calcium and magnesium ion solution (a mixture of calcium chloride and magnesium chloride), sediment suspension (particle size 0.01-0.1 mm), and rust particles (Fe2O3 content ≥95%) to the circulating water system to simulate water quality with different hardness and impurity contents. Water hardness is adjusted based on a concentration change function. To achieve precise control, for example, when the target hardness is 500 mg / L (calculated as CaCO3) and the initial hardness is 50 mg / L, k is set to 0.02 min. -1 (Calibrated using a system volume of 500L and an addition rate of 1L / min), this function allows the target concentration to be reached smoothly within 60 minutes, avoiding the impact of sudden changes in water quality on the heat exchanger. The water hardness adjustment range is 0-1000mg / L, and the impurity content adjustment range is 0-50g / m³, with real-time feedback on the adjustment effect via an online conductivity meter.

[0026] Temperature control unit: Utilizing a combination of an electric heating element (5kW) and a plate cooler, along with a Pt1000 platinum resistance temperature sensor (measurement accuracy ±0.05℃), closed-loop temperature control is achieved. The supply water temperature range is 30-95℃, and the return water temperature range is 20-70℃, with a temperature control accuracy of ±0.1℃. It can simulate temperature change curves during different stages of the heating system, including startup, stable operation, and shutdown.

[0027] Flow velocity control unit: Equipped with a variable frequency centrifugal pump (power 1.5kW, frequency range 0-50Hz), the flow velocity in the pipeline is fed back through an electromagnetic flow meter (measurement accuracy ±0.2%), with an adjustment range of 0.01-3m / s and a control accuracy of ±0.001m / s. It can simulate actual working conditions such as high flow velocity at the near end and low flow velocity at the far end of the pipeline network.

[0028] Battery power regulation unit: It adopts an adjustable DC power supply (output voltage 3.0-6.0V, accuracy ±0.01V), and connects to the battery interface of the tested household heat meter 300 through alligator clips. It simulates different power states such as new battery (6.0V), half-charge (4.5V), and low charge (3.3V), and collects battery voltage decay data in real time.

[0029] (II) Data Acquisition and Communication Module 200 The data acquisition and communication module 200 consists of 8 acquisition terminals, a data aggregation gateway, and a communication adapter module, supporting unified data acquisition from multiple brands and protocols of hot meters.

[0030] Communication Adaptation: Built-in configurable communication protocol library, covering mainstream protocols such as RS485, Modbus RTU, LoRa, and NB-IoT. For the proprietary protocols of common brands of heat meters such as Aimeidi, Aichuang, and Enleman, adaptation is achieved through protocol parsing scripts, without the need for additional modification of the heat meter.

[0031] Data Acquisition Parameters and Accuracy: Two types of data are acquired in real time: First, the metering response data of the tested household heat meter 300, including instantaneous flow rate (m³ / h), cumulative flow rate (m³), supply and return water temperature (°C), calculated heat (GJ), and instantaneous fault codes; second, operating condition parameter data, including pipeline pressure (MPa), standard meter flow rate (m³ / h), system temperature (°C), flow velocity (m / s), water conductivity (μS / cm), and battery voltage (V). The acquisition frequency can be configured within the range of 1-10Hz, with a default acquisition frequency of 5Hz. The data acquisition accuracy is ±0.001%FS, ensuring synchronization with standard metering data.

[0032] Data transmission: The acquisition terminal is connected to the data aggregation gateway via a shielded cable. The gateway uses an edge computing chip (ARM Cortex-A53) to perform preliminary filtering of the raw data (removing data that is obviously beyond the measurement range) and then transmits it to the data processing and storage module 400 via Ethernet or 4G module.

[0033] (III) Data Processing and Storage Module 400 The data processing and storage module 400 consists of an industrial server (CPU Intel Xeon E3-1230, memory 16GB, hard disk 4TB SSD) and data processing software. Its core functions include data storage, cleaning, alignment and standardization.

[0034] Data storage: The system uses a MySQL database to store both raw and processed data. It supports retrieval by heat table number, test time, operating condition type, and other dimensions. The data retention period is no less than 5 years, which meets the requirements for verification and traceability.

[0035] Data cleaning: For missing data, linear interpolation is used. Fill in; for abnormal data (such as instantaneous traffic surges of more than 10 times), use the 3σ criterion to identify and mark them, do not delete them directly, and preserve the integrity of the original data.

[0036] Data alignment and standardization: Based on the collection timestamp of the standard table, the metering data of the tested household heat meter 300 are time aligned (accuracy ±1ms); parameters of different magnitudes (such as temperature, flow rate, hardness) are standardized (normalized to the [0,1] interval) to provide a structured test dataset for the intelligent analysis core module 500.

[0037] (iv) Intelligent Analysis Core Module 500 The intelligent analysis core module 500 is the core unit of the system, integrating the working condition coupling influence analysis unit 510, the heat meter performance comprehensive evaluation unit 520, and the battery power decay prediction unit 530, and realizing multi-dimensional analysis based on the structured test dataset.

[0038] Operating condition coupling effect analysis unit 510: adopts a multivariate regression + SHAP hybrid model to quantify the impact of various operating condition parameters on the heat meter measurement error.

[0039] Step 1: Constructing Measurement Errors Operating parameters (Water hardness) Water supply temperature Flow rate Battery voltage The multivariate nonlinear regression model: ; Where the nonlinear basis function The regression coefficients are determined by fitting experimental data (e.g., power function for hardness, logarithmic function for temperature), and are solved using the least squares method (e.g., ...). , , wait).

[0040] Step 2: Input the trained regression model into the SHAP analysis tool to calculate the SHAP value for each parameter. For example, the average SHAP value of water hardness. Flow rate This indicates that water hardness has the greatest impact on measurement error.

[0041] Heat exchanger performance comprehensive evaluation unit 520: multi-dimensional evaluation based on entropy weight-TOPSIS method.

[0042] Step 1: Constructing the evaluation matrix ,in The number of heat exchangers being inspected. The evaluation indicators include average error, standard deviation of error, number of times the maximum permissible error exceeds the standard, and battery degradation rate.

[0043] Step 2: Calculate the entropy weight For example, the entropy value of the average error Weight Entropy value of battery degradation rate Weight .

[0044] Step 3: Calculate the weighted normalization matrix and Euclidean distance. , and relative closeness ,according to The performance levels were sorted into grades I-IV.

[0045] Battery capacity degradation prediction unit 530: Uses the ARIMA model to predict remaining service life (RUL).

[0046] For the collected battery voltage time series Perform a stationarity test (ADF test). If the stationarity is not normal, perform a first-order difference test. ); Determine model parameters using the AIC criterion. , Construct the ARIMA(2,1,1) model: ; Input historical voltage data (such as a continuous 72-hour voltage record) to predict future voltage decay trends. The predicted RUL (Relative Voltage Limit) is less than the safety threshold. Mark the battery risk at the end of the month.

[0047] (v) Calibration Optimization Recommendation Module 600 The calibration optimization suggestion module 600 generates dynamic calibration compensation coefficients based on the output of the operating condition coupling effect analysis unit 510. For example, regarding a certain brand of heat exchanger in high hardness ( mg / L), high temperature ℃, low flow rate The systematic positive error under the m / s condition is obtained by fitting the error sensitivity coefficient. , , Substitute into the formula: Calculations yielded This coefficient can be written into the heat meter calibration register to correct measurement errors.

[0048] (vi) Results output and data integration The system generates performance evaluation reports (including comprehensive performance ranking, operating condition impact patterns, and fault diagnosis results) and calibration optimization suggestions, which can be output in PDF, Excel, and Word formats. At the same time, it connects to the asset management system or meter procurement decision support system of heating companies through RESTful API and WebSocket interface to achieve real-time data sharing.

[0049] The workflow of this invention system is as follows: 1. Preset test plan: Set the operating parameters (such as water hardness 300mg / L, water supply temperature 60℃, flow rate 0.5m / s, battery voltage 4.5V), test duration (such as 24 hours), evaluation index weights, etc. through the host computer; 2. Working condition generation: The multi-parameter working condition simulation module 100 adjusts each parameter according to the preset scheme, and runs stably for 30 minutes after reaching the target working condition; 3. Data Acquisition: The data acquisition and communication module 200 synchronously acquires operating condition parameter data and metering response data of the household heat meter 300 under test, and transmits them to the data processing and storage module 400. 4. Data Processing: The data processing and storage module 400 completes data cleaning, alignment, and standardization to generate a structured test dataset; 5. Intelligent Analysis: The core intelligent analysis module 500 runs a hybrid model and outputs the influence law of metering error, the performance ranking of heat meter, and the battery RUL prediction results; 6. Calibration and Output: The calibration optimization suggestion module 600 generates dynamic calibration coefficients, and the result output module outputs a report and connects to the enterprise management system.

[0050] The following analysis will be conducted using three sets of examples: Example 1: Standard operating condition performance verification of residential heat meters of different brands 1. Test Object One DN20 household ultrasonic heat meter from each of three mainstream brands (Brand A: Aimeidi, Brand B: Aichuang, Brand C: Enleman) was selected, all of which were brand new and unused.

[0051] 2. Test conditions (standard conditions) Water hardness: 250 mg / L (CaCO3), impurity content: 5 g / m³; Supply water temperature: 50℃, return water temperature: 30℃, supply and return water temperature difference: 20℃; Flow velocity: 0.5 m / s (design rated flow velocity); Battery voltage: 6.0V (fully charged); Test duration: 24 hours, sampling frequency: 5Hz.

[0052] 3. Testing Procedures (1) Install the three household heat meters 300 to be tested on the test pipeline of the multi-parameter operating condition simulation module 100, and complete the communication protocol adaptation through the data acquisition and communication module 200 (Brand A adopts Modbus RTU protocol, Brand B adopts private RS485 protocol, and Brand C adopts LoRa protocol). (2) Start the multi-parameter operating condition simulation module 100, adjust the water quality, temperature, flow rate and battery voltage according to the standard operating condition parameters, and start timing after the operating condition stabilizes. (3) The data acquisition and communication module 200 synchronously acquires the metering response data (cumulative heat, instantaneous flow rate, supply and return water temperature, etc.) of the three heat meters and the standard meter data (high-precision ultrasonic heat meter, 0.2 grade), and transmits it to the data processing and storage module 400; (4) The data processing and storage module 400 completes data cleaning and standardization, and calculates the measurement error of each heat meter. ; (5) The operating condition coupling influence analysis unit 510 of the intelligent analysis core module 500 calculates the SHAP value of each parameter (under standard operating conditions). , , , ); (6) The comprehensive evaluation unit 520 for heat meter performance constructs an evaluation matrix and calculates the entropy weights (average error weight 0.36, error standard deviation weight 0.29, number of exceedances weight 0.15, battery degradation rate weight 0.20). The relative closeness is obtained by the TOPSIS method. (Level I) (Level II) (Level II); (7) The calibration optimization suggestion module 600 generates calibration coefficients based on the average error (1.2%) of brand C. ; (8) Output a performance evaluation report and connect it to the enterprise procurement decision system via API.

[0053] 4. Test Results like Figure 2 As shown, brand A heat meter has the best overall performance, with a measurement error of 0.5% and good stability; brand C heat meter has a slight positive error, which is reduced to 0.3% after calibration. The system recommends prioritizing brand A for procurement decisions.

[0054] Example 2: Test of heat meter metering performance and coupling effects under extreme operating conditions 1. Test Object One DN25 household mechanical heat meter (model X) of a certain brand has been in operation for 1 year (cumulative flow rate 500m³).

[0055] 2. Test conditions (extreme coupling conditions) Water hardness: 900 mg / L (high hardness), impurity content: 40 g / m³ (high impurities); Supply water temperature: 90℃ (high temperature), return water temperature: 65℃, supply and return water temperature difference: 25℃; Flow rate: 0.05 m / s (low flow rate, close to the minimum flow threshold). Battery voltage: 3.5V (low charge); Test duration: 48 hours, sampling frequency: 10Hz.

[0056] 3. Testing Procedures (1) The household heat meter 300 to be tested is installed in the test pipeline of the multi-parameter operating condition simulation module 100, and the data acquisition and communication module 200 establishes a connection through the Modbus RTU protocol; (2) Multi-parameter operating condition simulation module 100 is adjusted according to extreme operating conditions: water hardness is controlled by concentration change function. Adjustment was performed, and the target value was reached in 120 minutes; the temperature was gradually increased to 90℃, the flow rate was stabilized at 0.05m / s, and the battery voltage was set to 3.5V. (3) The data acquisition and communication module 200 continuously acquires data, focusing on recording the metering response delay and error fluctuation of the heat meter; (4) The data processing and storage module 400 identifies and marks 3 abnormal data points (instantaneous flow drop caused by impurities blocking the flow). (5) The measurement error is calculated by the working condition coupling influence analysis unit 510 of the intelligent analysis core module 500. SHAP values ​​of each parameter (Dominant influencing factors) , , This indicates that the coupling effect of high hardness, high impurities, and low flow rate is the main reason for the increased error. (6) The battery capacity degradation prediction unit 530 adopts the ARIMA(3,1,2) model. It inputs 48 hours of voltage data (3.5V to 3.3V) and predicts RUL=3 months, which is less than the safety threshold. Every month, the system generates an alert indicating that the battery needs to be replaced immediately; (7) Calibration optimization suggestion module 600 generates dynamic calibration coefficients. Meanwhile, it also proposed maintenance recommendations such as regularly cleaning meters and optimizing pipeline flow velocity ≥0.1m / s; (8) Output extreme working condition impact reports and fault early warning information, and connect to the enterprise operation and maintenance management system.

[0057] 4. Test Results like Figure 3 As shown, the metering error exceeded the standard under extreme coupling conditions. The allowable error is ≤2%, mainly due to the coupling effect of high hardness water and low flow rate. After calibration and battery replacement, the error was reduced to 0.8%, and the system provided precise optimization directions for operation and maintenance.

[0058] Example 3: Performance Ranking of Batch Heat Surface Meters and Battery Life Prediction 1. Test Object A heating company is looking to purchase 20 household ultrasonic heat meters of the same brand, from three production batches: batch 15, batch 28, and batches 3-7.

[0059] 2. Test conditions (simulating actual operating conditions) Water hardness: 400 mg / L, impurity content: 15 g / m³; Supply water temperature: 65℃, return water temperature: 40℃, supply and return water temperature difference: 25℃; Flow velocity: 0.3-1.0 m / s (dynamically changing, simulating flow velocity fluctuations in the pipeline network); Battery voltage: 5.0V (half charge); Test duration: 72 hours, sampling frequency: 8Hz.

[0060] 3. Testing Procedures (1) Install 20 household heat meters 300 to be tested in batch on the parallel test pipeline of the multi-parameter working condition simulation module 100 (supports simultaneous testing of up to 32 heat meters). The data acquisition and communication module 200 completes the communication connection of all heat meters within 30 minutes through the batch protocol adaptation function. (2) The multi-parameter working condition simulation module 100 operates according to the preset working conditions, and the flow rate is dynamically adjusted according to the cycle of 0.3→0.7→1.0→0.5→0.3m / s (one cycle every 12 hours); (3) The data acquisition and communication module 200 synchronously acquires the metering data and operating condition data of 20 heat meters and transmits them to the data processing and storage module 400. The storage capacity meets the data requirements of 72 hours × 8Hz × 20 heat meters. (4) The data processing and storage module 400 completes the cleaning and alignment of batch data and generates a 20-row × 4-column structured evaluation matrix; (5) The comprehensive evaluation unit 520 of the intelligent analysis core module 500 calculates the relative proximity of each heat meter. After sorting: 8 Class I heat meters ( (All are batch 2) 10 Class II heat meters ( (Batch 1: 3 pieces; Batch 3: 7 pieces) Two Class III heat meters ( All are batch 1). (6) The battery power degradation prediction unit 530 performed batch prediction on the battery voltage data of 20 heat meters and found that the RUL of the two Class III heat meters in batch 1 was 4 months (early warning), and the RUL of the remaining heat meters was ≥12 months. (7) The calibration optimization suggestion module 600 generates personalized calibration coefficients for two Class III heat meters. , ); (8) Output a batch heat meter performance ranking report and a battery life warning list, connect to the enterprise procurement and asset management system, and suggest purchasing heat meters in batch 2, and returning or calibrating the two heat meters in batch 1 for backup.

[0061] 4. Test Results like Figure 4 As shown, the system efficiently completed the verification of 20 batch heat meters, identified the best batch (batch 2) and unqualified heat meters, and achieved a 100% accuracy rate in battery life warning, providing data support for enterprise procurement cost control and asset optimization.

[0062] The intelligent analysis system for heat meter calibration data based on multi-parameter simulation of heating has the following advantages: First, this invention achieves accurate reproduction of multi-parameter coupled operating conditions in heating systems through a multi-parameter operating condition simulation module. It enables dynamic adjustment and precise control of key parameters such as water quality, temperature, flow rate, and battery charge. Combined with quantitative models such as concentration change functions, it can comprehensively simulate various real-world scenarios from standard to extreme operating conditions, making the verification process closer to the actual operating state of the heat meter. Simultaneously, relying on a multivariate regression and SHAP hybrid model, it not only quantifies the independent impact of each parameter on measurement error but also reveals the interactive coupling effect between parameters. This fills the technical gap in traditional verification where the laws governing operating conditions are unclear, providing more comprehensive and in-depth technical support for the analysis of heat meter measurement accuracy.

[0063] Secondly, this invention constructs a fully intelligent heat meter verification and evaluation system, significantly improving the scientific rigor of heat meter performance evaluation and the efficiency of operation and maintenance management. The multi-protocol adaptability of the data acquisition and communication module enables unified data acquisition and standardized analysis of heat meters from different brands and types, greatly reducing the operational complexity of multi-brand heat meter verification. The intelligent analysis core module integrates the entropy weight-TOPSIS comprehensive evaluation algorithm and the ARIMA battery life prediction model, which can output objective and fair heat meter performance rankings and accurate remaining service life warnings, solving the problems of traditional evaluation relying on manual experience and lagging fault diagnosis. The dynamic calibration compensation coefficient generated by the calibration optimization suggestion module can specifically correct systematic errors under specific operating conditions. Combined with early warning information and operation and maintenance suggestions, it forms full lifecycle management support for heat meters from procurement and selection, on-site calibration to replacement upon expiration, effectively reducing metering disputes and operation and maintenance costs caused by heat meter failures.

[0064] Third, this invention possesses strong industry applicability and promotional value, providing a powerful guarantee for energy conservation, emission reduction, and standardized development in the heating industry. The performance evaluation reports and operational condition impact analysis generated by the system seamlessly integrate with heating companies' asset management systems and procurement decision-making systems through standardized interfaces, enabling a data-driven decision-making model. This helps companies accurately select heat meter products with strong adaptability and high stability, optimizing procurement costs. Simultaneously, unified heat meter performance evaluation standards and verification procedures can standardize market access and quality assessment for heat meters, compelling companies to improve product quality and promoting industry technological upgrades. Furthermore, accurate metering calibration and fault early warning can reduce energy waste caused by heat metering deviations, improve the energy utilization efficiency of heating systems, align with the energy conservation and emission reduction development needs of the heating industry, and possess both economic value and social benefits.

[0065] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent analysis system for heat meter calibration data based on multi-parameter simulation of heating, characterized in that, include: The multi-parameter operating condition simulation module (100) is used to dynamically adjust and generate a multi-parameter coupled operating condition environment including water quality parameters, temperature parameters, flow rate parameters and battery power according to a preset test plan in a heating operating condition simulation laboratory. The data acquisition and communication module (200) is connected to the multi-parameter operating condition simulation module (100) and the household heat meter under test (300) for real-time acquisition of operating condition parameter data under the multi-parameter coupled operating condition environment and metering response data of the household heat meter under test (300). The data processing and storage module (400) is used to receive and store raw data from the data acquisition and communication module (200), and to clean, align and standardize the data to form a structured test dataset. The intelligent analysis core module (500) is connected to the data processing and storage module (400) and is used to calculate the performance indicators of the tested household heat meter (300) based on the structured test dataset, through the built-in analysis model and algorithm, analyze the multi-parameter coupling influence law, and generate a performance evaluation report. The intelligent analysis core module (500) includes at least an operating condition coupling influence analysis unit (510) and a heat meter performance comprehensive evaluation unit (520). The operating condition coupling effect analysis unit (510) adopts a hybrid model based on multivariate regression and Shapley additive interpretation (SHAP) to quantify the independent and interactive influence of each operating condition parameter on the heat meter measurement error. The heat meter performance comprehensive evaluation unit (520) constructs a multi-dimensional performance comprehensive evaluation model for heat meters based on the entropy weight-TOPSIS method, and outputs the comprehensive performance ranking and adaptability level of each tested heat meter.

2. The intelligent analysis system for heat meter calibration data based on multi-parameter simulation of heating as described in claim 1, characterized in that, The water quality parameter simulation in the multi-parameter operating condition simulation module (100) is achieved by adding a specific concentration of chemical substances to the circulating water. The concentration change function for simulating water hardness is as follows: ； in, express Ion concentration at time [time] This is the initial concentration. For the target concentration, This function represents a concentration adjustment coefficient related to the addition rate and system volume, enabling precise control and simulation of water quality changes.

3. The intelligent analysis system for heat meter calibration data based on multi-parameter simulation of heating as described in claim 1, characterized in that, The data acquisition and communication module (200) collects metering response data including instantaneous flow rate, cumulative flow rate, supply and return water temperature, calculated heat, and instantaneous fault code; and collects operating condition parameter data including pipeline pressure, standard meter flow rate, system temperature, flow rate, water conductivity, and battery voltage.

4. The intelligent analysis system for heat meter calibration data based on multi-parameter simulation of heating as described in claim 1, characterized in that, In the operating condition coupling impact analysis unit (510), the core algorithm of the model for quantifying the impact contribution is as follows: First, construct the measurement error. With multiple operating parameters Multivariate nonlinear regression model: ； in, For the intercept term, and For regression coefficients, and For nonlinear basis functions, This is the error term; Then, the trained regression model is used as the prediction model for SHAP analysis to calculate the parameters for each operating condition. SHAP value on each sample The mean of the SHAP values This indicates the average marginal contribution of the parameter to the measurement error.

5. The intelligent analysis system for heat meter calibration data based on multi-parameter simulation of heating as described in claim 1, characterized in that, In the comprehensive performance evaluation unit (520) of the heat exchanger, the specific implementation steps of the entropy weight-TOPSIS method are as follows: Step S1: Construct the evaluation matrix , line representative The evaluated hot list represents Several performance evaluation indicators, including average error, standard deviation of error, number of times the maximum permissible error exceeds the standard, and battery capacity decay rate under different operating conditions; Step S2: Calculate the objective weights of each evaluation index using the entropy weight method. : S21: Normalize the evaluation matrix to obtain the normalized matrix. For positive indicators For negative indicators ; S22: Calculate the... Entropy value of each indicator : ； in, ,like ,but , , ; S23: Calculate the... Weight of each indicator : ； Step S3: Calculate the comprehensive evaluation results using the TOPSIS method: S31: Construct a weighted normalization matrix , ; S32: Determine the ideal solution and negative ideal solution ,in , ; S33: Calculate the Euclidean distance from each heat exchanger to the positive and negative ideal solutions: ； ； S34: Calculate relative proximity : ； in accordance with The values ​​are used to rank all tested heat exchangers based on their overall performance.

6. The intelligent analysis system for heat meter calibration data based on multi-parameter simulation of heating as described in claim 1, characterized in that, The intelligent analysis core module (500) also includes a battery power degradation prediction unit (530). This unit predicts the remaining service life (RUL) of the hot meter battery based on the collected battery voltage time series data using a time series analysis algorithm. The prediction model adopts the differential autoregressive moving average (ARIMA) model, the general form of which is: ； in, for Battery voltage at any given time, For the shift operator ( ), Let be the difference order. It is a constant. These are the autoregressive coefficients. The moving average coefficient is... It is a white noise sequence. and These are the autoregression order and the moving average order, respectively.

7. The intelligent analysis system for heat meter calibration data based on multi-parameter simulation of heating as described in claim 6, characterized in that, The battery power degradation prediction unit (530) will predict the remaining service life. With a preset security threshold When comparing, When this happens, the system automatically generates a warning message and indicates that the heat meter has a potential battery power risk.

8. The intelligent analysis system for heat meter calibration data based on multi-parameter simulation of heating as described in claim 1, characterized in that, The system also includes a calibration optimization suggestion module (600), which generates dynamic calibration compensation coefficients based on the output of the operating condition coupling effect analysis unit (510) for systematic errors of a specific brand or model of heat meter under specific unfavorable operating conditions. The calculation formula is as follows: ； in, relative to standard temperature temperature difference, relative to standard flow rate The difference in flow velocity, This is the water hardness value, expressed as CaCO3, in units of... , This is the error sensitivity coefficient unique to this model of heat exchanger, obtained through fitting the influence analysis model.

9. The intelligent analysis system for heat meter calibration data based on multi-parameter simulation of heating as described in claim 1, characterized in that, The data acquisition and communication module (200) supports communication with multiple heat meters of different brands and with different communication protocols. It has a built-in configurable communication protocol library covering mainstream protocols such as RS485, Modbus RTU, LoRa, and NB-IoT. It achieves unified data acquisition and standardized parsing through protocol parsing and adaptation.

10. The intelligent analysis system for heat meter calibration data based on multi-parameter simulation of heating as described in any one of claims 1-9, characterized in that, The performance evaluation report and calibration optimization suggestions generated by the system are output to the asset management system or meter procurement decision support system of the heating company through a standardized data interface. The data interface supports RESTful API and WebSocket communication methods to realize data-driven decision support for heat meter procurement selection, replacement upon expiration, on-site calibration and operation and maintenance.

Images