A compensation device and method for improving the measurement accuracy of a direct-current charging pile

Abstract

The application discloses a compensation device and method for improving the measurement accuracy of a direct-current charging pile, relates to the technical field of direct-current charging piles, and comprises a measurement sampling module, a temperature compensation module, a ripple compensation module, a line voltage drop compensation module, a harmonic processing module, a main control module, a calibration module, an auxiliary module and a communication module. Each module cooperatively works, realizes the cooperative compensation of multiple error sources, improves the measurement accuracy of the direct-current charging pile, simultaneously reduces operation and maintenance costs, and improves the adaptability of the device.

Description

Technical Field

[0001] This invention relates to the field of DC charging pile technology, specifically to a compensation device and method for improving the metering accuracy of DC charging piles. Background Technology

[0002] The large-scale development of the electric vehicle industry has placed higher demands on the metering performance of DC charging piles. As the core basis for charging transaction settlement, the metering accuracy of DC charging piles not only affects the fairness of billing between operators and users, but also provides crucial data support for power grid-side power dispatch and load management. According to current standards, the metering accuracy of DC charging piles needs to reach 0.5 level or higher and maintain long-term stability under wide power ranges and complex environmental conditions. However, in practical engineering applications, the metering accuracy of DC charging piles is constrained by multiple factors, and existing technical solutions struggle to balance high accuracy with engineering adaptability, mainly exhibiting the following problems:

[0003] One issue is insufficient temperature adaptability. The operating environment inside DC charging piles experiences a wide temperature range (-35℃ to +70℃), causing significant drift in the resistance or gain of commonly used metering sensors (such as shunts and Hall elements) with temperature. Traditional single-sensor metering architectures lack systematic temperature control compensation, resulting in temperature drift errors of up to ±3%. While some solutions incorporate temperature calibration, these are mostly static lookup table methods, which are insufficient to address the dynamic errors caused by transient temperature rise lag during charging start-up and shutdown.

[0004] Secondly, there is the inaccuracy in ripple energy measurement. The AC / DC converter power supply module of a charging pile generates high-frequency ripple of 10-200kHz during operation. Existing metering methods often employ single-channel sampling with simple filtering, which either results in insufficient sampling rate leading to missed ripple energy measurements or fails to consider the phase difference between voltage and current ripples, introducing additional errors with an error margin exceeding 0.5%. Especially during transient processes such as charging start-up / stop and power switching, the ripple amplitude changes abruptly, causing the steady-state compensation model to fail, further exacerbating the measurement deviation.

[0005] Thirdly, line voltage drop introduces measurement errors. High-power DC charging piles (≥480kW) have long cables and charging currents up to 650A. The voltage drop caused by cable resistance and contact resistance can reach several volts. Traditional two-wire measurement methods cannot distinguish between load voltage drop and line loss, and include line loss in the measurement value, often resulting in an error exceeding 1%. Although some solutions attempt to optimize wiring, they do not fundamentally solve the measurement error problem caused by voltage drop.

[0006] Fourthly, harmonics and electromagnetic interference can lead to inflated metering results. Power grid harmonics, charger switching noise, and external electromagnetic interference couple into the sampling circuit. Traditional filtering methods can only suppress some low-frequency interference and are insufficient to effectively separate the fundamental and harmonic components, resulting in inflated metering results. Furthermore, existing solutions generally lack sampling data verification and fault tolerance mechanisms, making metering prone to failure when the sampling unit malfunctions.

[0007] Fifth, it suffers from poor long-term stability and high maintenance costs. Existing compensation models are mostly fixed parameter structures that are fixed before leaving the factory, making it difficult to adapt to factors such as component aging and environmental characteristic drift during long-term operation. The measurement accuracy gradually decreases over time. Calibration work requires disassembly, which involves a lot of manual intervention, is inefficient, and keeps maintenance costs high. Summary of the Invention

[0008] This invention provides a compensation device and method for improving the metering accuracy of DC charging piles, realizing collaborative compensation of multiple error sources, improving the metering accuracy of DC charging piles, reducing operation and maintenance costs, and enhancing the adaptability of the device.

[0009] This invention provides the following technical solution:

[0010] In a first aspect, the present invention discloses a compensation device for improving the metering accuracy of DC charging piles, comprising:

[0011] The metering sampling module adopts a four-wire measurement structure and is equipped with a dual-path synchronous sampling circuit, including a high-precision ADC main metering path for acquiring DC components and a high-speed sampling ADC ripple analysis path for acquiring ripple signals.

[0012] The temperature compensation module has a pre-stored temperature-error two-dimensional lookup table for multi-point calibration across the entire temperature range, and is equipped with a third-order transient thermal compensation unit to achieve millisecond-level predictive compensation for transient temperature rise or shrinkage during charging start-up and shutdown.

[0013] The ripple compensation module is used to extract the amplitude, frequency, phase, total harmonic distortion, ripple abrupt change coefficient, and decay time characteristic parameters of the ripple in real time, and calculate the ripple compensation energy based on the characteristic parameters and the least squares method online optimization model parameters.

[0014] The line voltage drop compensation module, based on the four-wire measurement structure, calculates the voltage drop caused by cable resistance and contact resistance in real time through the line impedance calculation model built into the main control module, and performs reverse correction on the measurement results.

[0015] The harmonic processing module has a built-in DSP digital signal processor. It separates the fundamental component and each harmonic component in voltage and current signals in real time through FFT harmonic analysis, and only performs integration measurement on the fundamental power.

[0016] The main control module is used to receive the sampling data and compensation parameters of each module, perform dynamic weight allocation of dual sensors and collaborative compensation calculation of multiple error sources, and synthesize the final measurement results.

[0017] The calibration module is used to periodically and automatically call the built-in standard source to perform self-testing on the metrology system and correct long-term drift errors.

[0018] Secondly, the present invention discloses a compensation method for improving the metering accuracy of DC charging piles. This method is applicable to the aforementioned compensation device and includes:

[0019] Step 1: The main control module starts the self-test of each module, loads the full temperature range temperature-error two-dimensional lookup table, piecewise linearization calibration parameters and dynamic ripple compensation basic model parameters, and completes the standard self-test;

[0020] Step 2: The metering sampling module collects the DC components of voltage and current through the main metering path, and collects the original voltage and current signals through the ripple analysis path, and transmits them synchronously to the main control module and the ripple compensation module.

[0021] Step 3: The temperature compensation module collects the sensor's operating temperature in real time. The main control module assigns sensor data weights based on the temperature value and performs basic temperature drift compensation by combining the temperature-error two-dimensional lookup table. If it is the charging start-stop stage, the third-order transient thermal compensation unit is activated to compensate for the error caused by the sensor's temperature rise lag or cold contraction in advance based on the thermal inertia prediction model.

[0022] Step 4: The main control module receives the current value of the current loop and the actual voltage value of the voltage sampling loop, and calculates the voltage drop caused by the cable resistance and contact resistance in real time through the built-in line impedance calculation model, and performs reverse correction on the voltage data after temperature compensation.

[0023] Step 5: The ripple compensation module performs high-frequency bandpass filtering on the original signal collected by the ripple analysis path to extract the pure ripple signal, and analyzes the characteristic parameters of the ripple through the FFT harmonic analysis unit; the main control module determines whether the charging condition is steady state or transient, calls the corresponding ripple compensation sub-model, optimizes the model parameters online using the least squares method, calculates the ripple compensation energy and superimposes it into the basic metering data;

[0024] Step 6: The harmonic processing module uses a DSP digital signal processor to perform FFT harmonic analysis on the pre-compensated voltage and current signals, separating the fundamental effective component from the ineffective components of each harmonic, and retaining only the fundamental component for power integration; at the same time, it activates a multi-stage filtering unit to suppress electromagnetic interference and switching noise, and automatically removes peak interference sampling points based on the 3σ criterion.

[0025] Step 7: The main control module calls the built-in piecewise linearization calibration logic to match the current voltage and current values ​​to the corresponding range segment, and performs nonlinear error correction on the multi-dimensional compensated measurement data through real-time table lookup and linear interpolation algorithm.

[0026] Step 8: The calibration module periodically starts the standard self-testing unit to correct long-term drift errors; if a remote calibration command is received from the cloud platform, the calibration coefficient is updated through the remote calibration unit; the main control module summarizes all compensation and calibration data and synthesizes the final power metering result.

[0027] Step 9: The communication module transmits the final measurement results to the charging pile main control board via the CAN bus, and simultaneously uploads the measurement data, calibration records and equipment operating status to the cloud platform via the network port.

[0028] Step 10: Repeat steps 2 to 9 throughout the entire charging process to dynamically adapt to changes in charging power and ambient temperature. After charging is completed, save complete metering and compensation records.

[0029] The present invention has the following beneficial effects:

[0030] 1. This invention adopts a collaborative temperature control scheme of dual sensors and third-order transient thermal compensation, which overcomes the shortcomings of traditional single-sensor temperature drift compensation. By combining temperature-error two-dimensional lookup table and transient thermal inertia prediction, it not only solves the steady-state temperature drift problem, but also offsets the transient temperature rise / cold contraction error during the charging start-up and shutdown phase, and the temperature drift suppression effect is greatly improved.

[0031] 2. This invention adopts a dual-path sampling and scenario-specific ripple compensation model. It uses a dual-path design of high-precision ADC and high-speed sampling ADC, combined with FFT harmonic analysis and least squares method online optimization, and constructs ripple compensation sub-models for steady state and transient state to accurately capture high-frequency ripple energy. This solves the problems of missed measurement, mismeasurement and sudden increase in transient error in traditional solutions, and significantly reduces ripple error.

[0032] 3. This invention adopts a four-wire system and an integrated line voltage drop compensation architecture, which completely separates the current loop from the voltage sampling loop. With the built-in integrated design, the sampling leads are shortened, and line voltage drop and lead interference are suppressed from the hardware source. Combined with real-time impedance calculation correction, line loss error is reduced, making it suitable for high-power supercharging scenarios. Attached Figure Description

[0033] Figure 1 This is a flowchart of the method in Embodiment 2 of the present invention.

[0034] Figure 2 This is a line graph comparing the measurement errors before and after compensation at different temperatures in Embodiment 3 of the present invention.

[0035] Figure 3 This is a line graph comparing the metering errors before and after compensation under different output currents in Embodiment 3 of the present invention.

[0036] Figure 4 This is a bar chart comparing the metering errors before and after compensation for different ripple and harmonic operating conditions in Embodiment 3 of the present invention.

[0037] Figure 5 This is a bar chart showing the measurement error of full-range segmented calibration in Embodiment 3 of the present invention.

[0038] Figure 6This is a bar chart comparing the measurement errors of composite working conditions before and after multi-algorithm fusion compensation in Embodiment 3 of the present invention. Detailed Implementation

[0039] The technical solutions of the embodiments of this specification will be explained and described below with reference to the accompanying drawings. However, the following embodiments are only preferred embodiments of this specification and not all of them. Other embodiments obtained by those skilled in the art based on the embodiments in the implementation methods without creative effort are all within the protection scope of this specification.

[0040] Example 1

[0041] This invention discloses a compensation device for improving the metering accuracy of DC charging piles, comprising:

[0042] The metering sampling module adopts a four-wire measurement structure and is equipped with a dual-path synchronous sampling circuit, including a high-precision ADC main metering path for acquiring DC components and a high-speed sampling ADC ripple analysis path for acquiring ripple signals.

[0043] Specifically, the metering sampling module adopts a four-wire (Kelvin) measurement structure, completely separating the current loop and the voltage sampling loop. The current loop uses a low-temperature drift manganese copper shunt and a digital Hall sensor to form a dual-sensor structure. The voltage sampling loop collects the actual voltage at the charging pile output terminal through a precision resistor voltage divider network to avoid line voltage drop interference. It is equipped with a high-precision ADC of 16 bits or more (main metering path) and a high-speed sampling ADC of 1MHz or more (ripple analysis path), and the voltage and current signals are collected synchronously through the dual paths, taking into account the acquisition accuracy of DC component and ripple signal. The shunt is selected as a low-temperature drift model with a temperature coefficient ≤5ppm / ℃, the Hall sensor is an anti-interference digital Hall sensor, and the resistor voltage divider network is composed of 0.01% precision resistors to ensure the basic accuracy of sampling.

[0044] The temperature compensation module has a pre-stored two-dimensional lookup table of temperature and error for multi-point calibration across the entire temperature range, and is equipped with a third-order transient thermal compensation unit to achieve millisecond-level predictive compensation for transient temperature rise or contraction during charging start-up and shutdown.

[0045] Specifically, the temperature compensation module has a built-in PT1000 temperature sensor that collects the operating temperature of the dual sensors (shunt and Hall sensor) in real time, completing a temperature sampling every 5ms. It sets a temperature threshold trigger unit to automatically switch the dual sensor operating mode according to the temperature range. At the same time, it pre-stores a temperature-error two-dimensional lookup table (LUT) with multi-point calibration across the entire temperature range (-40~85℃). In conjunction with the third-order transient thermal compensation unit, it achieves millisecond-level predictive compensation for transient temperature rise / contraction during charging start-up and shutdown, thus offsetting thermal inertia hysteresis error.

[0046] Among them, "automatic switching of dual sensor working modes according to temperature range" includes shunt-dominated mode, hybrid weighted mode, and Hall-dominated mode, which automatically switches or dynamically allocates weights according to temperature range to achieve high-precision measurement across the entire temperature range.

[0047] (1) Splitter-dominated mode (low temperature range: -40℃~-20℃):

[0048] The weighting of the manganese-copper shunt is 85%–100%.

[0049] The weight of digital Hall sensors ranges from 0% to 15%.

[0050] Compensation is based primarily on the shunt measurement value and secondarily on the Hall effect value.

[0051] (2) Mixed weighted mode (medium temperature range: -20℃~70℃):

[0052] The weight of the manganese-copper shunt linearly decreases from 85% to 0%, while the weight of the digital Hall sensor linearly increases from 15% to 100%.

[0053] The two measurement data are smoothly weighted and fused according to real-time temperature.

[0054] (3) Hall-dominated mode (high temperature zone: 70℃~85℃):

[0055] The digital Hall sensor has a weight of 100%, the manganese copper shunt has a weight of 0%, and the Hall sensor measurement value is used as the current measurement reference.

[0056] The temperature compensation module collects the operating temperature of the dual sensors in real time. The main control module assigns weights to the dual sensor data based on the temperature values ​​and performs basic temperature drift compensation by combining a temperature-error two-dimensional lookup table. At the same time, it determines the charging status (start / stop / steady state). If it is in steady state charging, it maintains dynamic weight allocation and basic temperature drift compensation to ensure that the temperature drift error is controlled within ±0.2%. If it is in the charging start / stop stage, it activates the third-order transient thermal compensation unit. Based on the thermal inertia prediction model, it compensates in advance for errors caused by sensor temperature rise lag or cold contraction, and outputs the basic voltage and current data after temperature compensation.

[0057] The third-order transient thermal compensation unit is specifically designed to address the temperature rise lag error caused by sensor thermal inertia under conditions of sudden large current changes during the charging start-stop phase. It achieves high-precision transient compensation through a three-level linkage algorithm, and the compensation steps are as follows:

[0058] (1) Start-up and stop condition identification: Real-time acquisition of charging current and calculation of current change rate dI / dt. When the current rises rapidly from 0 and dI / dt > 50A / s, it is determined to be the charging start condition; when the current falls rapidly back to 0 and dI / dt < -50A / s, it is determined to be the charging stop condition. After successful identification, the third-order transient thermal compensation mechanism is automatically started.

[0059] (2) First-order thermal inertia temperature prediction: Based on the first-order RC model of sensor thermal inertia, the temperature lag deviation is compensated, and the calculation formula is as follows:

[0060] ;

[0061] In the formula: Real-time temperature measurement using the PT1000 temperature sensor; To predict the actual junction temperature for the sensor; The factory-calibrated thermal time constant is 0.2–0.5 s. This represents the real-time temperature change rate.

[0062] (3) Second-order dynamic temperature drift error correction: Using the predicted junction temperature as an index, the pre-stored temperature-error two-dimensional lookup table (LUT) is retrieved to obtain the transient temperature drift error and correct the original current. The calculation formula is as follows:

[0063] ;

[0064] ;

[0065] In the formula: To fuse the raw current from the two sensors; This is for transient temperature drift error; This is the current after preliminary correction.

[0066] (4) Third-order smoothing fusion: A first-order low-pass filtering algorithm is used to smooth the corrected data to avoid measurement jitter caused by abrupt changes in compensation. The calculation formula is as follows:

[0067] ;

[0068] In the formula: This is a smoothing coefficient, with a value ranging from 0.1 to 0.3. This is the compensation current data from the previous sampling period.

[0069] (5) Output final compensation current: The compensated current data is sent to the main control module to participate in subsequent multi-dimensional compensation calculations.

[0070] The ripple compensation module is used to extract the amplitude, frequency, phase, total harmonic distortion, ripple abrupt change coefficient, and decay time characteristic parameters of the ripple in real time, and calculate the ripple compensation energy based on the characteristic parameters and the online optimization model parameters using the least squares method.

[0071] Specifically, the ripple compensation module includes a high-frequency bandpass filter unit, an FFT harmonic analysis unit, and a dynamic ripple compensation model unit. The high-frequency bandpass filter unit extracts the pure ripple signal from the ripple analysis path, while the FFT harmonic analysis unit extracts the amplitude, frequency, phase, total harmonic distortion, ripple mutation coefficient, decay time, and other characteristic parameters of the ripple in real time. The dynamic ripple compensation model unit optimizes the model parameters online based on the characteristic parameters and the least squares method, calculates the ripple compensation power in steady-state and transient scenarios, and integrates to obtain the ripple compensation energy. The transient scenario constructs a transient phased sub-model based on the ripple mutation coefficient and decay time to adapt to transient conditions such as charging start-stop and power switching.

[0072] The steady-state / transient condition determination mechanism is as follows:

[0073] Define ripple mutation coefficient This serves as the sole criterion for determining operating condition switching. In the formula: This represents the change in ripple current at the current moment. This represents the ripple current amplitude of the previous operating cycle.

[0074] Operating condition determination rules: When When the condition is determined to be steady-state, the steady-state ripple foundation compensation model is activated; when When the condition is determined to be a transient operating condition of charging start-stop or power jump, the transient phased attenuation sub-model is activated.

[0075] The high-frequency bandpass filter unit employs a high-frequency ripple component extraction algorithm to extract the pure ripple signal in the ripple analysis path, including:

[0076] The original sampled signal contained a mixed DC effective component, high-frequency ripple, and environmental noise. The pure high-frequency ripple signal was accurately separated through differential operations.

[0077] ;

[0078] In the formula: The purified high-frequency ripple voltage and current; This is the DC metering component after low-pass filtering; This is the raw sampled signal from the high-speed ADC.

[0079] Traditional metering methods calculate ripple power solely by multiplying voltage and current amplitudes, neglecting ripple phase difference, leading to inaccurate or missed ripple power measurements. This invention introduces ripple phase power factor correction to accurately calculate the true instantaneous ripple power. :

[0080] ;

[0081] In the formula: The real-time phase difference between voltage ripple and current ripple is obtained by frame-by-frame analysis using FFT.

[0082] Further establish dynamic compensation power models for different scenarios:

[0083] Under steady-state conditions, the ripple frequency and amplitude are stable, and accurate compensation is achieved using fixed fitting coefficients.

[0084] ;

[0085] In the formula: The steady-state ripple compensation coefficient is obtained from the factory calibration combined with least squares fitting.

[0086] Under transient start-stop conditions, the ripple exhibits rapid abrupt changes and exponential decay characteristics. A dynamic compensation model is constructed by introducing a time decay factor:

[0087] ;

[0088] In the formula: The ripple attenuation time constant; This is a transient dynamic correction coefficient, adapted to the sudden change error of ripple during start-up and shutdown.

[0089] Then, the least squares method was used to optimize the model parameters online.

[0090] To address model mismatch issues caused by component aging and long-term operating condition drift, this model undergoes online iterative optimization every 10ms, aiming to minimize the sum of squared residuals between the standard ripple power and the calculated power. The calculation formula is as follows:

[0091] Real-time updates through iteration The core compensation parameters enable adaptive calibration of the model, eliminating long-term metrological drift.

[0092] Then, integrate the ripple compensation power; the calculation formula is as follows:

[0093] Instantaneous compensation power cannot be directly used for electricity billing. Complete ripple compensation energy is obtained through time-domain integration and accumulation, continuously correcting the metering results. The calculation formula is as follows: ;

[0094] To adapt to the embedded engineering operation of the main control module (MCU), a discretized iterative integration method is adopted: In the formula: The ripple compensation calculation cycle is fixed at 10ms. Accumulate ripple compensation energy for the current moment; This is the accumulated compensation energy from the previous cycle.

[0095] Finally, the DC fundamental frequency energy and the high-frequency ripple compensation energy are fused to obtain the high-precision metered energy under all operating conditions. The calculation formula is as follows: In the formula: The precise DC fundamental frequency power after temperature and line voltage drop compensation; This is for dynamic compensation of electrical energy for high-frequency ripple.

[0096] The line voltage drop compensation module, based on the four-wire measurement structure, calculates the voltage drop caused by cable resistance and contact resistance in real time through the line impedance calculation model built into the main control module, and performs reverse correction on the measurement results.

[0097] Specifically, based on a four-wire measurement structure, the current value of the current loop and the actual voltage value of the voltage sampling loop are collected in real time. Through the line impedance calculation model built into the main control module, the voltage drop caused by cable resistance and contact resistance is calculated in real time, and the measurement results are corrected in reverse. At the same time, an integrated design is adopted to integrate the shunt, Hall sensor and meter, shorten the sampling line to the centimeter level, eliminate interference and voltage drop of long leads, and further reduce line error.

[0098] This module employs a dynamic adaptive line impedance calculation model, abandoning the traditional fixed resistance compensation method. It can adapt in real time to impedance drift caused by cable temperature rise, component aging, and joint oxidation and loosening, accurately calculating real-time line loss and voltage drop and performing reverse correction on the collected voltage data. Simultaneously, it adopts an integrated design, embedding the sensor and metering module within the module, shortening the sampling lead to less than 5cm, completely eliminating interference and additional voltage drop from long leads, and stably controlling the line voltage drop error to below 0.1%.

[0099] The line impedance calculation model distinguishes between the cable body impedance and the connector contact impedance. The formula for calculating the total real-time line impedance is as follows:

[0100] ;

[0101] In the formula: This represents the total real-time impedance of the line. Real-time resistance of the copper core cable body; This refers to the real-time contact resistance of terminals, wire harnesses, and connectors.

[0102] The cable's resistance dynamically drifts with operating temperature, and is corrected in real time using a copper resistance temperature characteristic model. The formula for calculating the total real-time impedance of the line is:

[0103] ;

[0104] In the formula: The nominal resistance of the cable is the factory-calibrated resistance at 20℃ (normal temperature). The temperature coefficient of the copper conductor is fixed at 0.00393 / ℃. This refers to the real-time operating temperature of the cable.

[0105] Contact resistance is susceptible to slow drift due to oxidation, loosening, and aging. The equipment undergoes real-time calibration and updates under no-load self-test conditions every morning. The formula for calculating the real-time resistance of the copper core cable is:

[0106] ;

[0107] In the formula: Provides a standard source to output the load reference voltage; The measured port voltage of the device; The standard constant current is used for self-testing.

[0108] Based on real-time line impedance, the instantaneous voltage drop loss of the line is accurately calculated using the following formula:

[0109] ;

[0110] In the formula: This refers to the real-time voltage drop of the line. The charging circuit samples the current in real time.

[0111] Finally, by reverse compensation, the true voltage at the load terminal is restored, and high-precision metered voltage data is obtained. The formula for calculating the true effective voltage of the load is:

[0112] ;

[0113] In the formula: The measured sampling voltage at the device port; This is the corrected actual effective voltage of the load used for electricity billing.

[0114] The harmonic processing module has a built-in DSP digital signal processor. It separates the fundamental component and each harmonic component in the voltage and current signals in real time through FFT harmonic analysis, and only performs integration measurement on the fundamental power.

[0115] Specifically, the harmonic processing module has a built-in DSP digital signal processor that uses FFT harmonic analysis to separate the fundamental wave (effective component) and each harmonic (ineffective component) in the voltage and current signals in real time. It only integrates and measures the fundamental power to eliminate the false power caused by harmonics. It is equipped with a multi-stage filtering unit, including hardware LC low-pass filter, software moving average filter and Kalman filter. Combined with the 3σ criterion, it automatically discards peak interference sampling points, comprehensively suppresses electromagnetic interference and switching noise, avoids false power, and improves the stability of the sampling signal.

[0116] The main control module is used to receive sampling data and compensation parameters from each module, perform dynamic weight allocation of dual sensors and collaborative compensation calculation of multiple error sources, and synthesize the final measurement results.

[0117] Specifically, the main control module uses a high-performance MCU to receive sampling data and compensation parameters from each module, perform dynamic weight allocation of dual sensors and collaborative compensation calculation of multiple error sources, and synthesize the final measurement result. It has a built-in full-range segmented linearization calibration logic, which divides the 0~1000V voltage and 0~650A current into 8~16 segments, and calibrates the nonlinear error of each segment independently. Through real-time table lookup and linear interpolation, it ensures the accuracy of the entire range.

[0118] The calibration module is used to periodically and automatically call the built-in standard source to perform self-testing on the metrology system and correct long-term drift errors.

[0119] Specifically, the calibration module includes a standard source self-testing unit and a remote calibration unit. The standard source self-testing unit automatically calls the built-in standard source periodically (daily / weekly) to perform self-testing on the metrology system and correct long-term drift errors. The remote calibration unit connects to the cloud platform via a communication module, supporting remote distribution of calibration coefficients from the cloud platform, achieving calibration without disassembling the device and reducing maintenance costs. This invention adopts a full-range segmented calibration and remote calibration mechanism without disassembly, independently calibrating nonlinear errors within each range to ensure accuracy across the entire range. By combining standard source self-testing with remote calibration via the cloud platform, it solves the problems of traditional model parameter fixation, poor long-term stability, and high calibration and maintenance costs, thereby improving equipment maintenance efficiency.

[0120] The communication module adopts a dual communication method of CAN bus and network port. On the one hand, it communicates with the main control board of the charging pile to output the final metering results and related metering data (charging time, DC component time series curve, ripple characteristic parameters, etc.); on the other hand, it communicates with the cloud platform to upload metering data, calibration records and equipment operating status, and supports remote monitoring and parameter debugging.

[0121] The auxiliary module includes a display screen, a keyboard, and a power supply module. The display screen is used to show measurement results, error data, and equipment status. The keyboard is used for on-site parameter setting and calibration operations. The power supply module uses a switching power supply with ±15V and 5V outputs to ensure stable power supply to each module.

[0122] Example 2

[0123] Please see Figure 1 As shown, this invention discloses a compensation method for improving the metering accuracy of DC charging piles. This method is applicable to the compensation device in Embodiment 1 and includes:

[0124] Step 1: Device initialization. The main control module starts self-test of each module, loads the full-temperature-error two-dimensional lookup table (LUT), piecewise linearization calibration parameters and dynamic ripple compensation basic model parameters, completes standard self-test, and ensures that each module is working properly.

[0125] Step 2: Dual-path synchronous sampling. The metering sampling module collects the DC components of voltage and current through the main metering path (low-pass filter and high-precision ADC), and collects the original voltage and current signals through the ripple analysis path (high-pass filter and high-speed sampling ADC), and transmits them synchronously to the main control module and the ripple compensation module.

[0126] Step 3: Temperature drift compensation. The temperature compensation module collects the operating temperature of the dual sensors in real time. The main control module assigns weights to the dual sensor data according to the temperature values ​​and performs basic temperature drift compensation by combining a temperature-error two-dimensional lookup table. At the same time, it determines the charging status (start / stop / steady state). If it is in the charging start / stop stage, the third-order transient thermal compensation unit is activated. Based on the thermal inertia prediction model, it compensates for the error caused by sensor temperature rise lag or cold contraction in advance and outputs the basic voltage and current data after temperature compensation. If it is in steady state charging, it maintains dynamic weight allocation and basic temperature drift compensation to ensure that the temperature drift error is controlled within ±0.2%.

[0127] Step 4: Line voltage drop compensation. The main control module receives the current value of the current loop and the actual voltage value of the voltage sampling loop transmitted by the metering sampling module. Through the built-in line impedance calculation model, it calculates the voltage drop value caused by cable resistance and contact resistance in real time, and performs reverse correction on the voltage data after temperature compensation to eliminate the metering deviation caused by line voltage drop. At the same time, relying on the hardware advantages of integrated design, it synchronously verifies the interference of sampling line leads to ensure that the line voltage drop error is reduced to below 0.1%.

[0128] Step 5: Ripple Compensation. The ripple compensation module performs high-frequency bandpass filtering on the raw signal acquired by the ripple analysis path to extract the pure ripple signal. The FFT harmonic analysis unit analyzes the characteristic parameters of the ripple, such as amplitude, frequency, phase, and abrupt change coefficient. The main control module determines the charging condition (steady-state / transient), calls the corresponding ripple compensation sub-model, optimizes the model parameters online using the least squares method, calculates the ripple compensation power, obtains the ripple compensation energy through integration, and superimposes it into the basic metering data to complete the ripple error compensation and solve the problems of missed or over-measured ripple energy.

[0129] Step 6: Harmonic and Interference Compensation. The harmonic processing module uses a DSP digital signal processor to perform FFT harmonic analysis on the pre-compensated voltage and current signals, separating the fundamental effective component from the ineffective components of each harmonic, and retaining only the fundamental component for power integration. At the same time, a multi-stage filtering unit is activated to suppress electromagnetic interference and switching noise through hardware LC low-pass filtering, software moving average filtering, and Kalman filtering in sequence. Combined with the 3σ criterion, peak interference sampling points are automatically eliminated to avoid false power increases caused by harmonics and interference.

[0130] Step 7: Full-range segmented calibration. The main control module calls the built-in segmented linearization calibration logic to match the current voltage and current values ​​to the corresponding range segments (0~1000V, 0~650A divided into 8~16 segments). Through real-time table lookup and linear interpolation algorithm, nonlinear error correction is performed on the measurement data after multi-dimensional compensation to ensure stable measurement accuracy across the entire range.

[0131] Step 8: Calibration and Result Synthesis. The calibration module periodically (daily / weekly) starts the standard self-test unit to perform self-test on the current metering system and correct drift errors caused by component aging during long-term operation. If a remote calibration command is received from the cloud platform, the calibration coefficient is updated through the remote calibration unit to complete calibration without disassembly. The main control module summarizes all compensation and calibration data, synthesizes the final power metering result, and synchronously stores the metering data, compensation parameters, and calibration records.

[0132] Step 9: Data transmission and status monitoring. The communication module transmits the final metering results, charging time, ripple characteristics, and other related data to the charging pile main control board via the CAN bus to support billing and settlement. At the same time, it uploads the metering data, calibration records, and equipment operating status to the cloud platform via the network port to support remote monitoring, parameter debugging, and fault diagnosis. The auxiliary module's display screen shows the metering results, error data, and equipment status in real time, facilitating on-site operation and maintenance.

[0133] Step 10: Cyclic execution and dynamic adaptation. Throughout the charging process, repeat steps 2 to 9 to collect, compensate, and calibrate data in real time, and dynamically adapt to operating conditions such as charging power switching and ambient temperature changes to ensure continuous and stable metering accuracy. After charging is completed, save complete metering and compensation records to complete one metering compensation process.

[0134] Example 3

[0135] I. Overall Experimental Testing Conditions

[0136] All experiments in this embodiment were conducted in strict accordance with the GB / T 17215-2022 DC power metering verification standard, with unified testing environment and electrical parameters to ensure that the experiments are reproducible and the data is traceable.

[0137] The overall testing conditions are as follows: voltage test range 200V~1000V, current test range 50A~650A, ambient temperature range -20℃~+60℃, ambient humidity controlled at 45%~75%RH (non-condensing); the system power supply is AC 380V±10%, 50Hz±1Hz clean mains power. The test conditions comprehensively cover normal temperature steady state, high and low temperature changes, long cable voltage drop, multi-level ripple and harmonic loads, and full-range segmented operating range, fully replicating various complex operating scenarios of commercial DC charging piles.

[0138] II. Experimental testing equipment and parameters of the device under test

[0139] 2.1 Parameters of the device under test

[0140] The device under test in this embodiment is a DC-1000-650-0.2S type high-precision power metering unit for DC charging piles. Its rated operating voltage range is 200V~1000V DC, and its rated operating current range is 50A~650A DC. Without algorithm compensation, its basic accuracy is 0.5S level. The device uses a 0.05-level high-precision shunt paired with a 16-bit synchronous sampling ADC to achieve high-speed and accurate acquisition of electrical parameters. Its operating temperature range is -20℃~+70℃, and it supports CAN2.0B bus data output, making it suitable for high-precision metering commercial scenarios for high-power DC charging piles.

[0141] 2.2 The standard test equipment parameter table is as follows:

[0142] DC standard power source ZY-1000-650 Level 0.02 Provides traceable standard voltage, current, and power output. DC standard energy meter SD-0.02 level Level 0.02 As a reference device for calculating measurement errors High and low temperature test chamber WGD / SJ-1000 Temperature control accuracy ±0.5℃ Simulates a working environment with a full temperature range of -20℃ to +60℃. Programmable DC electronic load IT8832B Level 0.05 Simulate different ripple and harmonic nonlinear load conditions High-precision data acquisition instrument Agilent 34970A Level 0.01 Voltage, current, and temperature data are collected and recorded simultaneously.

[0143] 2.3 Measurement Error Calculation Method

[0144] This embodiment uniformly adopts the industry-standard formula for calculating electricity metering errors. In the formula: The measured electrical energy value of the charging pile under test. This is the reference energy value for a standard energy meter.

[0145] III. Summary Table of Standardized Experimental Data

[0146] Table 1. Comparison of measurement errors before and after temperature drift compensation across the entire temperature range (corresponding to...) Figure 2 (Line graph comparing measurement errors before and after compensation at different temperatures)

[0147] -20℃ -0.75 ±0.18 0℃ -0.42 ±0.15 25℃(normal temperature) ±0.30 ±0.10 45℃ +0.45 ±0.16 60℃ +0.68 ±0.19

[0148] Table 2 Comparison of line voltage drop compensation errors under different cable operating conditions (corresponding to...) Figure 3 (Line graph comparing metering errors before and after compensation under different output currents)

[0149] 200A 5.0m -0.32 ≤0.07 350A 7.0m -0.41 ≤0.08 500A 9.0m -0.53 ≤0.09 650A 11.0m -0.64 ≤0.10

[0150] Table 3 Comparison of compensation accuracy under different ripple and harmonic operating conditions (corresponding to...) Figure 4 (Bar chart comparing metering errors before and after compensation for different ripple and harmonic operating conditions)

[0151] Low ripple conditions ±0.32 ≤0.12 Medium ripple condition ±0.55 ≤0.15 High ripple harmonic operating conditions ±0.82 ≤0.18

[0152] Table 4 Comparison of metrological accuracy under full-range segmented calibration (corresponding to...) Figure 5 (A bar chart of measurement error in segmented calibration across the full range)

[0153] Low voltage and low current range ±0.65 ≤0.15 Medium voltage and medium current range ±0.48 ≤0.12 High voltage and high current section ±0.58 ≤0.16

[0154] Table 5. Comparison of Comprehensive Working Condition Accuracy of Multi-Algorithm Integration and Collaborative Compensation (corresponding to...) Figure 6 (Bar chart comparing measurement errors under composite working conditions before and after multi-algorithm fusion compensation)

[0155] Mild coupling interference +0.72 ≤0.28 ≤0.11 Moderate coupling interference +0.85 ≤0.35 ≤0.14 Severe extreme coupling interference +0.93 ≤0.42 ≤0.17

[0156] IV. Specific Experimental Procedure

[0157] 4.1 Temperature Drift Compensation Experiment Across the Entire Temperature Range

[0158] 4.1.1 Test Principle: The electrical parameters of semiconductor devices such as metering sampling resistors, operational amplifiers, and ADC chips inside DC charging piles exhibit inherent drift characteristics with changes in ambient temperature. At low temperatures, the operating point of these devices shifts, easily generating negative metering errors; at high temperatures, increased leakage current and gain shift in these devices easily generate positive metering errors.

[0159] Traditional metering schemes use fixed sampling parameters and lack dynamic temperature correction capabilities, making them unsuitable for wide-temperature-range working scenarios. This results in large fluctuations in metering accuracy and poor stability due to seasonal temperature differences and outdoor high and low temperature environments.

[0160] This invention is equipped with a full-temperature-range dynamic temperature compensation algorithm, which can collect ambient temperature and device core temperature in real time, dynamically fit temperature drift correction curve, accurately offset the inherent temperature drift error of the device, and achieve stable and high-precision measurement across the entire temperature range.

[0161] 4.1.2 Test Conditions: Standard atmospheric pressure test environment, isolating external interference such as line voltage drop, load harmonic ripple, and range nonlinearity. A gradient constant temperature environment is constructed using a high-precision high and low temperature test chamber. All test equipment is preheated and calibrated. The device under test is placed at a constant temperature for 2 hours to ensure consistent internal and external temperatures and eliminate temperature hysteresis errors. A uniform 500V / 300A rated steady-state charging condition is adopted, and a single temperature gradient condition is used for continuous steady-state testing for 1 hour.

[0162] 4.1.3 Graded temperature operating parameters: Low temperature extreme condition (-20℃, simulating the extreme cold outdoor environment in winter), low temperature transition condition (0℃, simulating the transition environment between low temperature and normal temperature), normal temperature reference condition (25℃, standard calibration environment), high temperature transition condition (45℃, simulating the normal high temperature environment in summer), and high temperature extreme condition (60℃, simulating the extreme high temperature environment of outdoor exposure in summer).

[0163] 4.1.4 Complete traceability data and technical analysis: as shown in Table 1 and Figure 2 As shown, under the extreme low-temperature condition of -20℃, the standard electrical energy is 150.000 kWh, the metered value before compensation is 148.875 kWh (error -0.75%), and the metered value after compensation is 150.270 kWh (error +0.18%); under the low-temperature transition condition of 0℃, the standard electrical energy is 150.000 kWh, the metered value before compensation is 149.370 kWh (error -0.42%), and the metered value after compensation is 150.225 kWh (error +0.15%); under the standard temperature reference condition of 25℃, the standard electrical energy is 150.000 kWh, and the metered value before compensation is 150.450 kWh. kWh (error +0.30%), after compensation, the metered value is 150.150 kWh (error +0.10%); under the 45℃ high-temperature transition condition, the standard electrical energy is 150.000 kWh, the metered value before compensation is 150.675 kWh (error +0.45%), and the metered value after compensation is 150.240 kWh (error +0.16%); under the 60℃ high-temperature extreme condition, the standard electrical energy is 150.000 kWh, the metered value before compensation is 151.020 kWh (error +0.68%), and the metered value after compensation is 150.285 kWh (error +0.19%).

[0164] Technical Analysis: Traditional metering equipment exhibits significant temperature drift characteristics, with a maximum error approaching 0.8% across the entire temperature range. This makes it completely unsuitable for high-precision billing under extreme high and low temperature conditions. After correction using the dynamic temperature compensation algorithm of this invention, the metering error across all temperature gradient conditions converges to within ±0.2%, effectively solving the core problems of poor temperature adaptability and temperature drift inaccuracies in traditional charging piles, and significantly improving the metering stability of the equipment during all-season outdoor operation.

[0165] 4.2 Voltage Drop Compensation Experiment for Long Cable Lines

[0166] 4.2.1 Test Principle: Commercial DC charging piles commonly involve long cable runs in actual installation scenarios. The copper core cables have inherent line impedance, leading to significant voltage drops and power losses under high-current charging conditions. Traditional metering devices only collect output electrical parameters at the charging pile end and cannot identify cable transmission losses. The larger the output current and the longer the cable, the more severe the line loss error, continuously generating negative metering deviations, resulting in underestimation of billing and reduced site revenue. This invention's dynamic line voltage drop compensation algorithm can dynamically calculate real-time transmission losses based on output current, cable impedance, and cable length, adaptively adding a voltage drop correction coefficient to accurately compensate for metering deviations under high-current conditions with long cables.

[0167] 4.2.2 Test conditions: A clean test environment with a standard constant temperature of 25℃, no harmonic interference, and no temperature drift error. A fixed output voltage of 600V DC is used, matched with 120mm² national standard copper core cable commonly used in commercial charging piles. The test simulates actual engineering conditions with different wiring lengths and different output currents. All terminals are crimped and tightened, and the line contact impedance is standardized. The steady-state continuous test under a single set of working conditions is conducted for 1 hour. The equipment is preheated and calibrated in advance.

[0168] 4.2.3 Graded cable operating condition parameters: light current short cable operating condition (200A / 5m, simulating short-distance low-current charging scenario), medium-low current medium-long cable operating condition (350A / 7m, simulating conventional commercial cabling scenario), medium current long cable operating condition (500A / 9m, simulating medium-distance high-power charging scenario), and high current extreme long cable operating condition (650A / 11m, simulating long-distance, high-current extreme fast charging scenario).

[0169] 4.2.4 Complete traceability data and technical analysis: as shown in Table 2 and Figure 3 As shown, under the 200A / 5m operating condition, the standard electrical energy is 60.000kWh, the metered value before compensation is 59.808kWh (error -0.32%), and the metered value after compensation is 60.042kWh (error ≤0.07%); under the 350A / 7.0m operating condition, the standard electrical energy is 120.000kWh, the metered value before compensation is 119.508kWh (error -0.41%), and the metered value after compensation is 120.096kWh (error ≤0.08%). Under 500A / 9m operating conditions, the standard electrical energy is 180.000kWh, the metered value before compensation is 179.046kWh (error -0.53%), and the metered value after compensation is 180.162kWh (error ≤0.09%); under 650A / 11.0m operating conditions, the standard electrical energy is 240.000kWh, the metered value before compensation is 238.464kWh (error -0.64%), and the metered value after compensation is 240.240kWh (error ≤0.10%).

[0170] Technical Analysis: Traditional metering schemes cannot compensate for line transmission losses. The negative metering error deteriorates linearly with the increase of current and cable length, and the metering deviation is serious under high current and long cable conditions.

[0171] After correction by the dynamic voltage drop compensation algorithm of this invention, the metering error of all-gradient cables and all-current conditions is controlled within 0.10%, which completely solves the metering inaccuracy problem caused by transmission loss of long cables in charging piles and is perfectly adapted to various engineering wiring scenarios.

[0172] 4.3 Ripple and Harmonic Load Compensation Experiment

[0173] 4.3.1 Test Principle: When charging piles operate under nonlinear battery loads and rectifier loads, they generate ripple and multiple harmonics, causing ADC sampling distortion and mean shift, which are the core causes of dynamic measurement inaccuracies. Traditional hardware filtering cannot adapt to dynamic gradient interference, and the error continues to worsen as the disturbance intensity increases. This invention utilizes a self-developed graded ripple and harmonic suppression algorithm, which can identify the distortion level in real time and dynamically match the correction coefficient, achieving high-precision measurement under complex loads.

[0174] 4.3.2 Test conditions: ambient temperature 25℃, no external interference, clean mains power environment, uniform 500V / 300A steady-state condition, equipment preheating for 30 minutes and zero-point calibration completed, setting low, medium and high gradient load disturbances, single-condition steady-state test for 1 hour.

[0175] 4.3.3 Classification of operating parameters: Low ripple condition (ripple coefficient <1%, pure resistive load, no harmonics); Medium ripple condition (ripple coefficient 3%, superimposed with 3rd and 5th low-frequency harmonics, harmonic amplitude 5%); High ripple and harmonic condition (ripple coefficient 5%, superimposed with 3rd / 5th / 7th compound harmonics, harmonic amplitude 8%).

[0176] 4.3.4 Complete traceability data and technical analysis: as shown in Table 3 and Figure 4 As shown, under low ripple conditions, with a standard energy of 150,000 kWh, the error before compensation is ±0.32%, and the error after compensation is ≤0.12%; under medium ripple conditions, with a standard energy of 150,000 kWh, the error before compensation is ±0.55%, and the error after compensation is ≤0.15%; under high ripple and harmonic conditions, with a standard energy of 150,000 kWh, the error before compensation is ±0.82%, and the error after compensation is ≤0.18%.

[0177] Technical Analysis: Traditional metering lacks load disturbance suppression capabilities, and metering errors deteriorate significantly with increasing ripple and harmonic complexity, exceeding 0.8% under extreme operating conditions, failing to meet high-precision billing requirements. This invention's hierarchical adaptive compensation algorithm can accurately filter out interference from different load levels, correct distorted sampling data, and control errors under all gradient interference conditions to within 0.18%.

[0178] 4.4 Full-range segmented calibration experiment

[0179] 4.4.1 Test Principle: Analog devices such as metering chips and sampling resistors have inherent nonlinear characteristics. Low-voltage, small-signal devices are easily affected by background noise and zero-point offset, while high-voltage, high-current conditions are prone to device saturation distortion. Traditional single calibration coefficients cannot adapt to the full-range characteristics, resulting in large differences in accuracy under light load and full load, and poor consistency across the entire range. This invention adopts a full-range segmented calibration strategy, independently configuring compensation coefficients for low, medium, and high ranges to accurately correct nonlinear deviations in each interval.

[0180] 4.4.2 Test conditions: Standard ambient temperature environment, isolated from temperature, line and load interference, covering the full working range of charging piles from 200V to 1000V and 50A to 650A, three typical business conditions are set, the equipment is uniformly preheated and calibrated, and a single working condition steady state test is conducted for 1 hour to ensure that the data are comparable, accurate and effective.

[0181] 4.4.3 Graded range parameters: low voltage and low current 200V / 50A (light load charging scenario), medium voltage and medium current 500V / 300A (rated conventional charging scenario), high voltage and high current 1000V / 650A (extreme full load fast charging scenario).

[0182] 4.4.4 Complete traceability data and technical analysis: as shown in Table 4 and Figure 5 As shown, under low-voltage, low-current conditions, the standard electrical energy is 10.000 kWh, the error before compensation is ±0.65%, and the error after compensation is ≤0.15%; under medium-voltage, medium-current conditions, the standard electrical energy is 150.000 kWh, the error before compensation is ±0.48%, and the error after compensation is ≤0.12%; under high-voltage, high-current conditions, the standard electrical energy is 650.000 kWh, the error before compensation is ±0.58%, and the error after compensation is ≤0.16%.

[0183] Technical Analysis: Traditional metering devices suffer from inherent nonlinearity defects. A single calibration coefficient cannot adapt to the full-range characteristics, resulting in significant issues such as zero-point offset with small signals and saturation distortion in high-current devices, leading to extremely poor accuracy consistency across the entire range. This invention corrects nonlinear errors in each interval through segmented and zoned independent calibration and dynamic matching compensation coefficients. The metering error for all operating conditions across the entire range converges to within ±0.2%, achieving high-precision metering across the entire range.

[0184] 4.5 Comprehensive Verification Experiment of Multi-Algorithm Fusion and Collaborative Compensation

[0185] 4.5.1 Test Principle: The aforementioned individual experiments all adopted a single-variable test method to independently verify the correction performance of each individual compensation algorithm, which can accurately verify the independent optimization effect of each technical module. However, in the actual commercial operation scenario of DC charging piles, temperature changes, cable transmission losses, load ripple harmonic interference, and device range nonlinearity errors do not exist independently. Multiple types of errors will couple and superimpose with each other, causing a significant decrease in measurement accuracy. A single compensation algorithm is difficult to take into account the comprehensive error problem caused by compound interference.

[0186] To fully verify the overall adaptability and comprehensive metering performance of the multi-algorithm integrated collaborative compensation architecture of this invention, this experiment constructs a complex working environment with multiple variables coupled. It enables a complete set of fusion algorithms, including transient temperature compensation, dynamic correction of line voltage drop, graded suppression of ripple and harmonics, and full-range segmented calibration. The metering accuracy differences between no compensation, single-algorithm compensation, and multi-algorithm fusion compensation are compared to verify the comprehensive correction capability of the system under complex working conditions.

[0187] 4.5.2 Test Conditions: The experiment adopts a multivariate coupled superposition test mode, abandoning single-variable constraints and maximizing the reproduction of complex on-site working conditions. Three sets of gradient composite interference conditions are set up to cover light, medium and heavy actual application scenarios, fully matching the complex operating states of commercial charging piles in all scenarios. All experimental equipment is preheated for 30 minutes and zero-point calibration is completed. Each set of conditions is subjected to constant temperature and steady-state continuous testing for 1 hour.

[0188] 4.5.3 Graded Comprehensive Operating Condition Parameters:

[0189] Operating Condition 1: Mild coupling interference (35℃ medium temperature + 7m cable + low ripple and no harmonics + rated medium range).

[0190] Operating Condition 2: Moderate coupling interference (45℃ high temperature + 9m cable + medium ripple low frequency harmonics + large current range).

[0191] Operating Condition 3: Severe extreme coupling interference (60℃ extreme high temperature + 11m long cable + high ripple composite harmonics + full load and large range).

[0192] 4.5.4 Complete traceability data and technical analysis: as shown in Table 5 and Figure 6 As shown,

[0193] Under operating condition 1 with mild coupling interference, standard electrical energy 120.000 kWh, the uncompensated metering error is +0.72%, the optimal error with a single compensation algorithm is 0.28%, and the error converges to ≤0.11% after enabling multi-algorithm fusion compensation.

[0194] Under operating condition 2 with moderate coupling interference, standard electrical energy 150.000 kWh, the uncompensated metering error is +0.85%, the optimal error with a single compensation algorithm is 0.35%, and the error after multi-algorithm fusion compensation is ≤0.14%.

[0195] Under extreme coupling interference in the working condition, with a standard electrical energy of 180.000kWh, the uncompensated metering error is as high as +0.93%. The superposition of multiple errors leads to serious inaccuracies in traditional metering. The best compensation error of a single algorithm is only 0.42%, which cannot meet the high-precision metering standard. After compensation by a complete set of fusion algorithms, the metering error is reduced to ≤0.17%.

[0196] Technical Analysis: Under complex operating conditions with multi-dimensional error coupling and superposition, traditional uncompensated metering schemes exhibit extremely large errors. A single compensation algorithm can only correct a single-dimensional error and cannot offset composite metering deviations; even after compensation, the accuracy still cannot reach the 0.2S level high-precision standard. This invention's unique multi-algorithm integrated collaborative compensation mechanism can identify multiple types of coupled error sources in real time and dynamically allocate the correction weights of each compensation algorithm. This achieves synchronous full-domain correction of temperature drift, line loss, load harmonic interference, and range nonlinearity errors. The algorithms collaborate and complement each other, avoiding the limitations of single compensation techniques. Even under extremely harsh composite interference conditions, the system metering error can still be stably controlled within ±0.2%, fully verifying the universality, operational stability, and engineering practical value of this invention's technical solution.

[0197] This invention proposes a multi-dimensional collaborative compensation scheme that combines transient temperature compensation, dynamic correction of line voltage drop, graded suppression of ripple and harmonics, and segmented calibration across the entire measurement range. This scheme can specifically eliminate four core measurement errors. Verified through multiple sets of single-variable control experiments and composite operating condition experiments, this invention enables charging piles to maintain stable measurement errors within ±0.2% under extreme and complex operating conditions, including full temperature range, full power range, long cable transmission, and strong load interference, fully meeting the 0.2S-level DC high-precision measurement industry standard.

[0198] This invention effectively solves the industry pain points of traditional charging piles, such as large temperature drift error, inaccurate loss of long cables, weak load anti-interference, and inconsistent accuracy across the entire range. It significantly improves the metering accuracy, environmental adaptability, and robustness of the equipment under working conditions, with outstanding technical advantages and extremely high engineering practicality and market promotion value.

[0199] The embodiments described above are merely preferred embodiments of this specification and are not intended to limit the scope of this specification. Any modifications and improvements made by those skilled in the art to the technical solutions of this specification without departing from the spirit of this specification should fall within the protection scope defined by the claims of this specification.

Claims

1. A compensation device for improving the measurement accuracy of a direct current charging pile, characterized in that, include: The metering sampling module adopts a four-wire measurement structure and is equipped with a dual-path synchronous sampling circuit, including a high-precision ADC main metering path for acquiring DC components and a high-speed sampling ADC ripple analysis path for acquiring ripple signals. The temperature compensation module has a pre-stored temperature-error two-dimensional lookup table for multi-point calibration across the entire temperature range, and is equipped with a third-order transient thermal compensation unit to achieve millisecond-level predictive compensation for transient temperature rise or shrinkage during charging start-up and shutdown. The ripple compensation module is used to extract the amplitude, frequency, phase, total harmonic distortion, ripple abrupt change coefficient, and decay time characteristic parameters of the ripple in real time, and calculate the ripple compensation energy based on the characteristic parameters and the least squares method online optimization model parameters. The line voltage drop compensation module, based on the four-wire measurement structure, calculates the voltage drop caused by cable resistance and contact resistance in real time through the line impedance calculation model built into the main control module, and performs reverse correction on the measurement results. The harmonic processing module has a built-in DSP digital signal processor. It separates the fundamental component and each harmonic component in voltage and current signals in real time through FFT harmonic analysis, and only performs integration measurement on the fundamental power. The main control module is used to receive the sampling data and compensation parameters of each module, perform dynamic weight allocation of dual sensors and collaborative compensation calculation of multiple error sources, and synthesize the final measurement results. The calibration module is used to periodically and automatically call the built-in standard source to perform self-testing on the metrology system and correct long-term drift errors.

2. The compensation device for improving the measurement accuracy of a direct current charging pile according to claim 1, characterized in that, In the metering and sampling module, the current loop adopts a dual-sensor structure consisting of a low-temperature drifting manganese copper shunt and a digital Hall sensor, while the voltage sampling loop collects the actual voltage at the output of the charging pile through a precision resistor voltage divider network to avoid line voltage drop interference.

3. The compensation device for improving the metering accuracy of a direct current charging pile according to claim 1, characterized in that, Under steady-state conditions, the temperature compensation module completes basic temperature drift compensation only through a two-dimensional temperature-error lookup table; under start-stop transient conditions, the third-order transient thermal compensation unit is activated to achieve high-precision temperature compensation across the entire temperature range and all operating conditions. The steps for current compensation in the third-order transient thermal compensation unit include: The charging current is collected in real time and the rate of change of current is calculated. Based on this, the charging start-up condition or the charging stop condition is determined. After successful identification, the third-order transient thermal compensation mechanism is automatically activated. Based on the first-order RC model of sensor thermal inertia, the temperature hysteresis bias is compensated to obtain the sensor's predicted true junction temperature. Using the predicted actual junction temperature as an index, a pre-stored temperature-error two-dimensional lookup table (LUT) is retrieved to obtain the transient temperature drift error and correct the original current, thus obtaining the preliminary corrected current. A first-order low-pass filter algorithm is used to smooth the preliminary corrected current data to avoid metering jitter caused by compensation abrupt changes; The final compensation current is output, and the compensated current data is sent to the main control module to participate in subsequent multi-dimensional compensation calculations.

4. The compensation device for improving the measurement accuracy of a direct current charging pile according to claim 1, characterized in that, The compensation steps of the ripple compensation module include: The original signal is acquired by a high-speed ADC, and the pure high-frequency ripple component is separated by high-frequency bandpass filtering to eliminate DC bias and low-frequency interference. Then, the core characteristic parameters such as ripple amplitude, frequency, phase, and abrupt change coefficient are analyzed frame by frame using FFT; the steady-state and transient compensation sub-models are automatically switched according to the ripple abrupt change coefficient, and the model parameters are optimized online using the least squares method to calculate the ripple compensation power; Finally, the ripple compensation energy is accumulated by the time-domain integration algorithm and superimposed on the DC fundamental energy to obtain high-precision metering energy under all operating conditions, thus completing high-precision ripple compensation.

5. The compensation device for improving the measurement accuracy of a direct current charging pile according to claim 4, characterized in that, The calculation of ripple compensation power includes: Under steady-state conditions, the ripple frequency and amplitude are stable. Accurate compensation is achieved using fixed fitting coefficients, and the calculation formula is as follows: ; wherein is a steady state ripple compensation coefficient; is a ripple instantaneous power; Under transient start-stop conditions, the ripple exhibits rapid abrupt changes and exponential decay characteristics. A dynamic compensation model is constructed by introducing a time decay factor, and the calculation formula is as follows: ; wherein is a ripple decay time constant; is a transient dynamic correction factor; is a ripple instantaneous power.

6. The compensation device for improving the metering accuracy of DC charging piles according to claim 5, characterized in that, The ripple compensation energy is obtained by integrating the ripple compensation power in the time domain. The calculation formula is as follows: ;in Power for ripple compensation; For the operating conditions of running in an MCU embedded project, a discretized iterative integration method is used for calculation, and the calculation formula is as follows: ; wherein is a ripple compensation operation period, fixed value 10 ms; is the current time accumulated ripple compensation energy; is the last period accumulated compensation energy.

7. The compensation device for improving the measurement accuracy of a direct current charging pile according to claim 1, characterized in that, The voltage drop compensation steps of the line voltage drop compensation module are as follows: The resistance of the cable body is calculated using a copper resistance temperature characteristic model. The calculation formula is: ;in The nominal resistance of the cable is the factory-calibrated resistance at 20℃ (normal temperature). The temperature coefficient of the copper conductor is fixed at 0.00393 / ℃. Real-time operating temperature of the cable; Calculate the real-time contact resistance of terminals, wire harnesses, and connectors. The calculation formula is: ;in Provides a standard source to output the load reference voltage; The measured port voltage of the device; The self-test standard constant current; calculating the total real-time impedance of the circuit , the formula is: ; Based on the total real-time impedance of the circuit, the instantaneous loss voltage drop of the circuit is calculated, and the calculation formula is: ; wherein is the real-time loss voltage drop of the circuit; is the real-time sampling current of the charging circuit; By restoring the true voltage at the load terminal through reverse compensation, the metered voltage data is obtained. The calculation formula is as follows: ;in The measured sampling voltage at the device port; This is the corrected actual effective voltage of the load used for electricity billing.

8. The compensation device for improving the measurement accuracy of a direct current charging pile according to claim 1, characterized in that, It also includes a communication module, which uses both CAN bus and Ethernet port for communication; It communicates with the main control board of the charging pile to output the final metering results and associated metering data. It communicates with the cloud platform to upload measurement data, calibration records, and equipment operating status, and supports remote monitoring and parameter debugging.

9. The compensation device for improving the measurement accuracy of a direct current charging pile according to claim 1, characterized in that, It also includes auxiliary modules, a display screen for showing measurement results, error data and equipment status, a keyboard for on-site parameter setting and calibration operations, and a power supply module that uses ±15V and 5V output switching power supplies to ensure stable power supply to each module.

10. A compensation method for improving the metering accuracy of a direct current charging pile, the method being applicable to the compensation device of any one of claims 1-9, characterized in that, include: Step 1: The main control module starts the self-test of each module, loads the full temperature range temperature-error two-dimensional lookup table, piecewise linearization calibration parameters and dynamic ripple compensation basic model parameters, and completes the standard self-test; Step 2: The metering sampling module collects the DC components of voltage and current through the main metering path, and collects the original voltage and current signals through the ripple analysis path, and transmits them synchronously to the main control module and the ripple compensation module. Step 3: The temperature compensation module collects the sensor's operating temperature in real time. The main control module assigns sensor data weights based on the temperature value and performs basic temperature drift compensation by combining the temperature-error two-dimensional lookup table. If it is the charging start-stop stage, the third-order transient thermal compensation unit is activated to compensate for the error caused by the sensor's temperature rise lag or cold contraction in advance based on the thermal inertia prediction model. Step 4: The main control module receives the current value of the current loop and the actual voltage value of the voltage sampling loop, and calculates the voltage drop caused by the cable resistance and contact resistance in real time through the built-in line impedance calculation model, and performs reverse correction on the voltage data after temperature compensation. Step 5: The ripple compensation module performs high-frequency bandpass filtering on the original signal collected by the ripple analysis path to extract the pure ripple signal, and analyzes the characteristic parameters of the ripple through the FFT harmonic analysis unit; the main control module determines whether the charging condition is steady state or transient, calls the corresponding ripple compensation sub-model, optimizes the model parameters online using the least squares method, calculates the ripple compensation energy and superimposes it into the basic metering data; Step 6: The harmonic processing module uses a DSP digital signal processor to perform FFT harmonic analysis on the pre-compensated voltage and current signals, separating the fundamental effective component from the ineffective components of each harmonic, and retaining only the fundamental component for power integration; at the same time, it activates a multi-stage filtering unit to suppress electromagnetic interference and switching noise, and automatically removes peak interference sampling points based on the 3σ criterion. Step 7: The main control module calls the built-in piecewise linearization calibration logic to match the current voltage and current values ​​to the corresponding range segment, and performs nonlinear error correction on the multi-dimensional compensated measurement data through real-time table lookup and linear interpolation algorithm. Step 8: The calibration module periodically starts the standard self-testing unit to correct long-term drift errors; if a remote calibration command is received from the cloud platform, the calibration coefficient is updated through the remote calibration unit; the main control module summarizes all compensation and calibration data and synthesizes the final power metering result. Step 9: The communication module transmits the final measurement results to the charging pile main control board via the CAN bus, and simultaneously uploads the measurement data, calibration records and equipment operating status to the cloud platform via the network port. Step 10: Repeat steps 2 to 9 throughout the entire charging process to dynamically adapt to changes in charging power and ambient temperature. After charging is completed, save complete metering and compensation records.

Images