DC system switch state intelligent detection system based on multi-point current sampling

Abstract

The invention discloses a DC system switch state intelligent detection system based on multi-point current sampling, and the system comprises a current sampling unit which is used for collecting the current data of a bus tie switch main loop of a DC system; the current sampling unit comprises micro-current sensor arrays which are arranged at an input end P1, an internal contact position P2 and an output end P3 of the bus tie switch and are used for detecting real-time current values of two sides and a middle node of the switch respectively; the signal conditioning and acquisition unit is connected with the current sampling unit; according to the method, the micro-current sensor arrays are deployed at a plurality of key positions of the main loop of the bus tie switch, current data of two sides and middle nodes of the switch are acquired, and the actual conduction state of the main loop is directly judged by using differential characteristics and correlation analysis of multi-point current. Even if the auxiliary contact loses efficacy, the system can still accurately identify the switch state based on the real distribution of the main loop current, and the dependence on the auxiliary contact is thoroughly eliminated.

Description

Technical Field

[0001] This invention relates to the field of DC system switch status detection, and in particular to an intelligent DC system switch status detection system based on multi-point current sampling. Background Technology

[0002] The DC system in a substation is a crucial infrastructure for ensuring the safe and stable operation of the power system. Its reliable operation directly affects the normal functioning of secondary equipment such as protection, control, and signaling systems within the substation. However, existing DC system switch status detection technologies have the following shortcomings: Traditional DC system switch status detection mainly relies on auxiliary contact signals, which has inherent flaws. The auxiliary contacts are mechanically independent of the main contacts; when the main contacts fail due to poor contact, ablation, or spring fatigue, the auxiliary contacts may still close normally, leading the control system to misjudge the actual switch status.

[0003] Some existing detection devices use a single-point current detection method, installing a current sensor on one side of the bus tie switch to determine the switch status by detecting the current flow. However, this method is easily affected by factors such as circulating current, charging current, and load fluctuations. When the battery pack is undergoing equalization or float charging, even if the bus tie switch is open, current still flows in the charging circuit, and single-point detection cannot distinguish whether this is normal charging current or circulating current after the bus tie switch is closed.

[0004] Traditional detection devices generally use fixed threshold criteria, which cannot adapt to the detection needs of different capacity levels and operating conditions. DC systems have different rated voltages, such as 110V and 220V, and battery capacities range from hundreds to thousands of ampere-hours, with a wide range of rated currents for switches. Using a uniform current threshold for state judgment may be insufficiently sensitive for large-capacity systems, while being overly sensitive for small-capacity systems. Furthermore, the internal resistance characteristics of batteries vary significantly under different charge / discharge states and ambient temperatures, and fixed thresholds cannot be dynamically adjusted to reflect these changes.

[0005] Existing detection devices lack the capability to monitor the insulation status and loop faults of DC systems. When a ground fault occurs in a DC system, the faulty loop cannot be located, requiring maintenance personnel to disconnect each feeder one by one for investigation, leading to widespread load loss and affecting system reliability. Simultaneously, potential circulating current issues in the DC system cannot be effectively identified, resulting in uneven discharge of battery banks and shortening their lifespan. Summary of the Invention

[0006] Therefore, the technical problem to be solved by the present invention is that the existing DC system switch status detection technology is insufficient.

[0007] The above-mentioned technical problems are solved by the following technical solution: This invention proposes an intelligent detection system for the switch status of a DC system based on multi-point current sampling, comprising: a current sampling unit for collecting current data of the main circuit of the DC system bus tie switch; the current sampling unit includes a micro-current sensor array deployed at the input terminal P1, internal contact position P2, and output terminal P3 of the bus tie switch, for detecting the real-time current values ​​at both sides and the middle node of the switch respectively; a signal conditioning and acquisition unit connected to the current sampling unit for filtering and analog-to-digital conversion of the collected current signal; and a main control processing unit connected to the signal conditioning and acquisition unit; the main control processing unit is configured to: acquire the current data of each sampling point, calculate the current difference characteristics and current correlation characteristics between adjacent sampling points, and combine them with preset state judgment logic to determine the actual on or off state of the bus tie switch.

[0008] The sensor at the input terminal P1 is located at the connection point connecting the first battery pack to the input terminal of the bus tie switch; the sensor at the output terminal P3 is located at the connection point connecting the output terminal of the bus tie switch to the second battery pack; and the sensor at the internal contact location P2 is located near the internal contact of the switch. These three sampling points are connected in series in spatial topology, and the sensors in the microcurrent sensor array are Hall current sensors.

[0009] In a preferred embodiment of the intelligent detection system for DC system switching status based on multi-point current sampling described in this invention: it further includes a threshold adaptive unit for dynamically adjusting the judgment threshold in the status judgment logic; when the device starts up, it identifies the capacity level of the DC system based on the monitored average floating charge current, and initializes the current difference threshold and correlation coefficient threshold according to the capacity level; during device operation, it statistically analyzes the current difference component distribution characteristics under historical normal closed states, calculates its mean and standard deviation, and updates the dynamic current difference threshold in real time.

[0010] In a preferred embodiment of the intelligent detection system for DC system switching state based on multi-point current sampling described in this invention: when extracting features, the main control processing unit needs to calculate the first current difference component Δ between the input terminal P1 and the internal contact position P2. I 12 And the second current difference component Δ between the internal contact position P2 and the output terminal P3. I 23 ; Calculate the first Pearson correlation coefficient r between input terminal P1 and internal contact position P2 within a set time window. 12 And the second Pearson correlation coefficient r between the internal contact position P2 and the output terminal P3. 23；The state judgment logic is as follows: when the current difference component is less than the dynamic current difference threshold and the correlation coefficient is greater than the correlation coefficient value, the switch is determined to be closed; when the current difference component is greater than the dynamic current difference threshold or the correlation coefficient is less than the correlation coefficient value, the switch is determined to be open.

[0011] In a preferred embodiment of the intelligent detection system for DC system switching status based on multi-point current sampling described in this invention: the threshold adaptive unit further includes a temperature compensation module; the device is equipped with a temperature sensor, and the temperature compensation module corrects the current difference threshold using a temperature coefficient based on the collected ambient temperature, with the reference temperature set to 25 degrees Celsius.

[0012] In a preferred embodiment of the intelligent detection system for DC system switch status based on multi-point current sampling described in this invention: it further includes an insulation monitoring unit and a transfer switch unit; the current sampling unit further includes feeder current sensors F1 to Fn installed at the outgoing ends of each DC feeder; the insulation monitoring unit uses the AC injection method to inject a low-frequency AC signal into the DC bus, and uses the feeder current sensors to detect the AC response current of each feeder to calculate the insulation resistance to ground; the transfer switch unit includes a main bus switch group and a backup bus switch group; when the main control processing unit receives a signal that the insulation resistance of a certain feeder is lower than the alarm threshold, it controls the transfer switch unit to seamlessly switch the faulty feeder from the main bus to the backup bus.

[0013] In a preferred embodiment of the intelligent detection system for DC system switch status based on multi-point current sampling described in this invention: the main control processing unit is further configured to perform a DC loop detection function, collect the output current of the two sets of batteries and the load current of all feeders; calculate the difference between the total output current of the batteries and the total load current of the feeders; when the absolute value of the difference continuously exceeds the circulating current judgment threshold, it is determined that there is a circulating current in the system, and the circulating current path is located by combining the current direction information of the internal contact position P2 of the bus tie switch.

[0014] In a preferred embodiment of the intelligent detection system for DC system switch status based on multi-point current sampling described in this invention: the main control processing unit is further configured with a status confirmation delay mechanism: when a change in switch status is detected, it continuously observes for a preset time length, and only when the judgment results of all sampling points are consistent within this time period is the switch status output confirmed and updated to filter out transient interference.

[0015] In a preferred embodiment of the intelligent detection system for the switching state of a DC system based on multi-point current sampling described in this invention: the signal conditioning and acquisition unit performs first-order low-pass filtering on the signals of each channel, and the filter cutoff frequency is set to 100Hz to attenuate high-frequency noise interference; the sampling frequency of the current sampling unit is set to 1kHz.

[0016] In a preferred embodiment of the intelligent detection system for DC system switch status based on multi-point current sampling described in this invention, a display alarm unit is also included. The display alarm unit adopts a TFT touch screen and uses different colors to indicate the switch closed, open, and pending confirmation states. The alarm function is divided into three levels: information level, warning level, and fault level, which correspond to different display colors and sound and light prompt strategies, respectively.

[0017] The beneficial effects of this invention are as follows: By deploying a micro-current sensor array at multiple key locations in the main circuit of the bus tie switch, current data from both sides of the switch and the intermediate node are collected. Utilizing the differential characteristics and correlation analysis of the multi-point currents, the actual conduction state of the main circuit can be directly determined. Even if the auxiliary contacts fail, the system can still accurately identify the switch state based on the true distribution of the main circuit current, completely eliminating the reliance on auxiliary contacts. A comprehensive analysis model for multi-point current sampling data is established. By calculating the current difference, current gradient, and correlation coefficient between adjacent sampling points, switch state characteristics are extracted. Utilizing the spatial distribution information and temporal correlation of multi-point data, the interference of bus tie circulating current, charging current, and load fluctuations on the detection results can be effectively distinguished. When charging current occurs, the current direction at each sampling point is consistent and the values ​​are close; when the bus tie is closed and circulating current is generated, the sampling points on both sides of the switch show obvious current differences; when the load is switched on and off, the current change exhibits transient characteristics. Multi-dimensional feature identification improves detection accuracy. An adaptive threshold adjustment algorithm was developed to calculate the optimal judgment threshold in real time based on parameters such as battery capacity, current charging and discharging state, historical operating data, and ambient temperature. The algorithm automatically identifies the system capacity level when the device starts up and continuously monitors changes in battery internal resistance and load characteristics during operation. It dynamically adjusts the current differential threshold and response delay. For different operating conditions, the threshold can be automatically optimized within a set range, so that the detection device always works in the best sensitivity state. This ensures detection accuracy under low current and avoids false triggering of large-capacity systems, achieving universal adaptation to various DC systems. The device integrates DC feeder insulation monitoring function. By installing insulation detection modules in each feeder circuit, it monitors the insulation resistance of each branch to ground in real time. When the insulation of a feeder is detected to be reduced, the device uses the transfer switch function to isolate the circuit from the main system to the backup bus. This ensures that the load of the branch does not lose power and facilitates online inspection of grounding points by operation and maintenance personnel, thus improving the efficiency of fault location. Add a DC loop detection function to identify whether there is abnormal circulating current in the system by analyzing the direction and amplitude relationship of the current at each sampling point. When circulating current is detected, the circulating current path is located by combining the switch status information, which guides the operation and maintenance personnel to adjust the operation mode, avoid uneven discharge and energy loss of the battery pack, extend the battery life and improve the system operating efficiency. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings of the embodiments of the present invention will be briefly described below. Obviously, the drawings described below only relate to some embodiments of the present invention and are not intended to limit the present invention. Wherein: Figure 1 A flowchart of the present invention is shown. Detailed Implementation

[0019] To enable those skilled in the art to better understand the present invention, the present invention will be further described in detail below with reference to specific embodiments and accompanying drawings.

[0020] The terminology used in this invention is that which is currently widely used in the art in consideration of the function of the invention; however, these terms may vary according to the intent of those skilled in the art, precedent, or new technology in the art. Furthermore, specific terms may be chosen by the applicant, and in such cases, their detailed meanings will be described in the detailed description of the invention. Therefore, the terms used in this specification should not be construed as simple names, but rather based on their meanings and the overall description of the invention.

[0021] Reference Figure 1 This embodiment provides an intelligent detection system for the switch status of a DC system based on multi-point current sampling, including: a current sampling unit for collecting current data of the main circuit of the DC system bus tie switch; the current sampling unit includes an array of micro-current sensors deployed at the input terminal P1, internal contact position P2, and output terminal P3 of the bus tie switch for detecting real-time current values ​​at both sides and the middle node of the switch respectively; a signal conditioning and acquisition unit connected to the current sampling unit for filtering and analog-to-digital conversion of the collected current signal; and a main control processing unit connected to the signal conditioning and acquisition unit; the main control processing unit is configured to: acquire current data at each sampling point, calculate the current difference characteristics and current correlation characteristics between adjacent sampling points, and combine them with preset state judgment logic to determine the actual on or off state of the bus tie switch.

[0022] The sensor at the input terminal P1 is located at the connection point connecting the first battery pack to the input terminal of the bus tie switch, the sensor at the output terminal P3 is located at the connection point connecting the output terminal of the bus tie switch to the second battery pack, and the sensor at the internal contact location P2 is located near the internal contact of the switch; these three sampling points are connected in series in spatial topology, and the sensors in the micro-current sensor array are Hall current sensors.

[0023] It also includes a threshold adaptive unit, which is used to dynamically adjust the judgment threshold in the state judgment logic; when the device starts up, it identifies the capacity level of the DC system based on the average value of the monitored floating charge current, and initializes the current difference threshold and correlation coefficient threshold according to the capacity level; during the operation of the device, it statistically analyzes the distribution characteristics of the current difference component under the historical normal closed state, calculates its mean and standard deviation, and updates the dynamic current difference threshold in real time.

[0024] When extracting features, the main control processing unit needs to calculate the first current difference component Δ between the input terminal P1 and the internal contact position P2. I 12 And the second current difference component Δ between the internal contact position P2 and the output terminal P3. I 23 ; Calculate the first Pearson correlation coefficient r between input terminal P1 and internal contact position P2 within a set time window. 12 And the second Pearson correlation coefficient r between the internal contact position P2 and the output terminal P3. 23； The status judgment logic is as follows: when the current difference component is less than the dynamic current difference threshold and the correlation coefficient is greater than the correlation coefficient value, the switch is determined to be closed; when the current difference component is greater than the dynamic current difference threshold or the correlation coefficient is less than the correlation coefficient value, the switch is determined to be open.

[0025] The threshold adaptive unit also includes a temperature compensation module; the device is equipped with a temperature sensor, and the temperature compensation module corrects the current differential threshold using a temperature coefficient based on the collected ambient temperature, with the reference temperature set at 25 degrees Celsius.

[0026] It also includes an insulation monitoring unit and a transfer switch unit; the current sampling unit also includes feeder current sensors F1 to Fn installed at the outgoing ends of each DC feeder; the insulation monitoring unit uses the AC injection method to inject a low-frequency AC signal into the DC bus and uses the feeder current sensors to detect the AC response current of each feeder to calculate the insulation resistance to ground; the transfer switch unit includes a main bus switch group and a backup bus switch group; when the main control processing unit receives a signal that the insulation resistance of a feeder is lower than the alarm threshold, it controls the transfer switch unit to seamlessly switch the faulty feeder from the main bus to the backup bus.

[0027] The main control processing unit is also configured to perform DC loop detection function, collect the output current of the two sets of batteries and the load current of all feeders; calculate the difference between the total output current of the batteries and the total load current of the feeders; when the absolute value of the difference continues to exceed the circulating current judgment threshold, it is determined that there is circulating current in the system, and the circulating current path is located by combining the current direction information of the internal contact position P2 of the bus tie switch.

[0028] The main control processing unit is also equipped with a status confirmation delay mechanism: when a change in the switch status is detected, it continuously observes for a preset time length. Only when the judgment results of all sampling points are consistent within this time period will the switch status output be confirmed and updated to filter out transient interference.

[0029] The signal conditioning and acquisition unit performs first-order low-pass filtering on the signals of each channel, with the filter cutoff frequency set to 100Hz to attenuate high-frequency noise interference; the sampling frequency of the current sampling unit is set to 1kHz.

[0030] It also includes a display warning unit, which uses a TFT touch screen to indicate the switch's closed, open, and pending confirmation status with different colors. The alarm function is divided into three levels: information level, warning level, and fault level, each corresponding to different display colors and sound and light prompt strategies.

[0031] The main circuit of the bus tie switch is equipped with three main sampling points, located at the switch input, switch output, and internal contact positions, respectively. The input sampling point, designated P1, is used to detect the current on the A-group side of the battery; the output sampling point, designated P3, is used to detect the current on the B-group side of the battery; and the internal contact sampling point, designated P2, is used to detect the actual current flow within the switch. Each sampling point is equipped with a Hall current sensor, with a measurement range of -500 amperes to +500 amperes, a resolution of 10 mA, and a response time of less than 5 milliseconds.

[0032] The sensor at sampling point P1 passes through the copper busbar connecting the positive terminal of battery group A to the input terminal of the bus tie switch. The sensor's output voltage signal is proportional to the current flowing through the copper busbar. The signal line is connected to input channel 1 of the signal conditioning unit via a shielded twisted pair cable. The sensor at sampling point P3 passes through the copper busbar connecting the output terminal of the bus tie switch to the positive terminal of battery group B, and its signal line is connected to input channel 3. The sensor at sampling point P2 is installed near the internal contacts of the switch. Due to space constraints, a miniature Hall sensor is used, and its signal line is led out to input channel 2 via a sealed connector. The three sampling points are connected in series in space, satisfying the analytical conditions of Kirchhoff's current law.

[0033] In addition to the main sampling point, the device adds feeder current sampling points, denoted as F1 to Fn, at the outgoing end of each DC feeder, where n is the total number of feeders, typically 8 to 16. Each feeder sampling point is also equipped with a Hall current sensor for monitoring the load current of each branch and detecting circulating current. The signal lines of the feeder sampling points are connected to the extended sampling channel for synchronous data acquisition with the main sampling point.

[0034] The sampling frequency is set to 1000 times per second, i.e., the sampling period is 1 millisecond. The sensor output voltage range at each sampling point is -5V to +5V, corresponding to a current measurement range of -500A to +500A. The voltage-current conversion relationship of the Hall sensor is as follows: In the formula, Let be the output voltage of the p-th sensor, in volts; The conversion factor of the Hall sensor is 0.01 volts per ampere; The current passing through the sensor is measured in amperes; p takes values ​​of 1, 2, and 3, corresponding to sampling points P1, P2, and P3, respectively.

[0035] The signal conditioning unit performs first-order low-pass filtering on each channel signal. The filter cutoff frequency is 100 Hz to attenuate high-frequency noise interference. The difference equation of the transfer function of the first-order low-pass filter in the discrete domain is: In the formula, The filtered voltage of the q-th channel at the k-th sampling time is expressed in volts. The original sensor voltage of the q-th channel at the k-th sampling time is expressed in volts. is the filter coefficient, with a value of 0.159, corresponding to a cutoff frequency of 100 Hz; k is the sampling number; q takes values ​​of 1, 2, and 3, corresponding to sampling points P1, P2, and P3 respectively.

[0036] The filtered signal is fed into a 16-bit analog-to-digital converter (ADC) with a conversion accuracy of 0.15 millivolts, corresponding to a current resolution of 15 milliamps. The relationship between the ADC's digital output value and the input voltage is as follows: In the formula, The digitized output value of the q-th channel at the k-th sampling time is a dimensionless integer ranging from 0 to 65535. The filtered voltage of the q-th channel at the k-th sampling time is expressed in volts. The lower limit of the reference voltage is set at -5 volts. The upper limit of the reference voltage is set to +5 volts; q takes values ​​of 1, 2, and 3, corresponding to sampling points P1, P2, and P3, respectively. The reference voltage of the analog-to-digital converter uses a high-precision reference source with a temperature drift of less than 10 ppm per degree Celsius, ensuring long-term measurement stability.

[0037] The data acquisition module records the current values ​​of the three main sampling points as follows: , , The unit is amperes. The current value at the feeder sampling point is denoted as... to The sampling time is recorded as... Where k is the sampling sequence number. Each sampling yields a current data set, which is stored in a circular buffer. The buffer depth is set to 1000, retaining the sampling data from the most recent second for subsequent analysis. The sampling time information is provided by the real-time clock of the main control unit, with a clock accuracy of 1 millisecond, synchronized with the sampling period. All sampling channels use a synchronous triggering method to ensure that the current values ​​at each point at the same time strictly correspond, avoiding analysis errors caused by time deviations.

[0038] The insulation monitoring unit uses AC injection to perform insulation testing on each feeder circuit. This unit includes a low-frequency AC signal generator, an injection coupling circuit, a response current detection circuit, and an insulation resistance calculation module. The signal generator produces a 1 Hz sinusoidal AC voltage signal with an amplitude of 5 volts, which is injected into the positive and negative buses of the DC system through a coupling transformer. The injection point is selected at the bus junction before the bus tie switch, ensuring that the AC signal covers all feeder circuits.

[0039] The time-domain expression of the AC injection signal is: In the formula, The instantaneous value of the injected AC voltage, in volts; The peak value of the injection voltage is 5 volts. The frequency of the injected signal is 1 Hz; t is the time variable, in seconds.

[0040] Each feeder is equipped with an AC response current sensor at its outgoing terminal to detect the branch's response current to an AC injection signal. The sensor uses a through-core current transformer, passing through the positive conductor of the feeder, which can separate the DC and AC components, outputting only the AC response current signal. The amplitude of the response current is inversely proportional to the insulation resistance of the feeder to ground; the lower the insulation resistance, the larger the response current.

[0041] The main control unit collects the AC response current amplitude of each feeder and records it as follows: to The unit is milliampere (mA). Based on the peak value of the injected signal voltage. Based on the response current, calculate the insulation resistance to ground of each feeder. The calculation formula is: In the formula, is the insulation resistance to ground of the i-th feeder, in kiloohms; The peak value of the injected AC signal voltage is in volts and is set to 5 volts. The AC response current amplitude of the i-th feeder is expressed in milliamperes; i is the feeder number, ranging from 1 to n. Under normal circumstances, the insulation resistance to ground of each feeder in the DC system should be greater than 50 kΩ. The device sets the insulation alarm threshold to 20 kΩ. When the calculated insulation resistance of a feeder is lower than this threshold, it is determined that there is a risk of insulation degradation in that circuit, triggering an insulation alarm.

[0042] During insulation testing, the injected signal amplitude and frequency remain constant to avoid interference with DC load equipment. Since the injection frequency is 1 Hz, far below the operating frequency of the DC load equipment, and the amplitude is only 5 volts, it will not affect the normal operation of the system. The response current detection uses a bandpass filter with a center frequency of 1 Hz and a bandwidth of 0.2 Hz, effectively suppressing noise interference at other frequencies and improving detection accuracy.

[0043] When the insulation monitoring unit detects that the insulation resistance of a feeder is lower than the alarm threshold, it transmits the faulty feeder number and insulation resistance value to the main control unit. The main control unit immediately initiates the third step of the loop isolation and transfer process to prevent the insulation fault from further deteriorating and causing the system to ground.

[0044] The transfer switch unit includes a main bus switch group and a standby bus switch group, with one transfer switch for each feeder. The main bus switch connects the feeder to the main DC bus, and the standby bus switch connects the feeder to the standby DC bus. The standby bus is powered by an independent battery bank with a capacity equivalent to the main system, and is normally in hot standby mode. During normal operation, the main bus switches for all feeders are closed, the standby bus switches are open, and the feeders draw power from the main bus.

[0045] When the insulation monitoring in the second step detects a decrease in the insulation of a feeder, the main control unit sends a transfer command to the transfer switch unit of that feeder. The transfer process adopts a "close first, then disconnect" method to ensure that the load does not lose power. The specific steps are as follows: First, close the standby bus switch of the feeder, so that the feeder is simultaneously powered by the main bus and the standby bus for 50 milliseconds to ensure power continuity; then open the main bus switch of the feeder to complete the circuit transfer. At this time, the feeder is only powered by the standby bus and is electrically isolated from the main system.

[0046] The transfer switch employs a fast vacuum switch with a closing and opening time of less than 20 milliseconds, ensuring high operational reliability. The switch status is dually confirmed via auxiliary contacts and multi-point current sampling (a feature of this patent) to prevent malfunctions. During the transfer process, the main control unit continuously monitors the load current of the feeder, confirming continuous and uninterrupted current flow to verify a successful transfer. After the transfer is complete, the insulation fault of the feeder is isolated in the backup system, the insulation level of the main system returns to normal, and other feeders remain unaffected and continue normal operation.

[0047] The isolated faulty feeder continues to be powered by the backup bus, and the operation of the load equipment is unaffected. Maintenance personnel can conduct a detailed online inspection of the circuit, using an insulation resistance tester to test each section and locate the specific grounding point. Because the faulty circuit is isolated from the main system, even if a short-term power outage is required for insulation testing during the inspection process, it will not affect the main system or other loads, greatly improving the safety and flexibility of fault handling.

[0048] The transfer switch unit also features both manual and automatic operating modes. In automatic mode, the device automatically performs the transfer based on insulation monitoring results; in manual mode, maintenance personnel can manually control the transfer switch via the touchscreen display of the alarm unit or the remote communication interface for planned maintenance or testing.

[0049] DC loop detection is based on the principle of energy conservation and multi-point current sampling data. Under normal operating conditions, the sum of the output currents of the two battery banks should equal the sum of the load currents of all feeders, satisfying power balance. When circulating current exists in the system, some current circulates between the battery banks without passing through the load, causing the sum of the battery output currents to exceed the sum of the load currents, resulting in energy loss.

[0050] The main control unit collects the output current of battery group A. and the output current of battery group B and all feeder load currents to Calculate the difference between the total output current of the battery and the total load current of the feeder, as the basis for determining circulating current: In the formula, This is the circulating current, measured in amperes. This is the output current of battery group A, in amperes. This is the output current of battery group B, measured in amperes. Let be the load current of the i-th feeder, in amperes; n be the total number of feeders; and i be the feeder number, ranging from 1 to n. Theoretically, under no-circulating current conditions, It should be close to zero. However, in actual operation, due to measurement errors and system losses, a small deviation is allowed.

[0051] The device is set to a circulating current detection threshold of 5 amperes. When the current remains within 10 consecutive seconds... When the absolute values ​​of all values ​​are greater than 5 amperes, circulating current is determined to exist in the system. Further analysis of the current at sampling point P2 inside the bus tie switch is needed. When the bus tie switch is closed and direction and or Conversely, the circulating current path is confirmed to be from battery group A to battery group B via the bus tie switch, or vice versa. This circulating current is usually caused by an imbalance in the voltage of the two battery groups, with the side with the higher voltage discharging to the side with the lower voltage.

[0052] Circulating current detection also incorporates the current direction information acquired in the first step. Each Hall current sensor can distinguish the current direction, outputting a positive voltage to indicate forward current and a negative voltage to indicate reverse current. By analyzing the combinations of current directions at each sampling point, the circulating current path can be accurately located. For example, when... For positive negative For positive and When the value exceeds the threshold, it indicates that battery group A is discharging to battery group B through the bus tie switch, forming a circulating current; when and All are positive but Approaching zero and When the value is close to zero, it indicates that the two sets of batteries are powered independently and there is no circulating current.

[0053] Upon detecting circulating current, the device sends a circulating current alarm to maintenance personnel via the display alarm unit, showing the magnitude and path of the circulating current. Maintenance personnel are advised to check the terminal voltages of both battery banks, adjust charging parameters or equalize the voltage, and eliminate the source of the circulating current.

[0054] The main control processing unit preprocesses the acquired raw current data. First, it calculates the moving average of the current at the three main sampling points to eliminate random noise and impulse interference. The moving window length is set to 10 sampling points, corresponding to a 10-millisecond time window. The moving average current at sampling point P1 is calculated as the arithmetic mean of the most recent 10 sampled currents. In the formula, Let P1 be the moving average current at the k-th sampling time, in amperes. This is the length of the sliding window, with a value of 10. Let be the original current value of sampling point P1 at the km-th sampling time, in amperes; m be the offset within the window, ranging from 0 to 9; and k be the sampling sequence number. Similarly, calculate the moving average current of sampling points P2 and P3. and The moving average method preserves the trend of current variation while effectively suppressing high-frequency noise, providing a stable data foundation for subsequent feature extraction.

[0055] Based on the current value after moving average, the current difference component between adjacent sampling points is calculated to identify the current distribution characteristics on both sides of the switch. The current difference component between input terminal P1 and internal contact P2 is denoted as... The calculation method is as follows minus The current difference between internal contact P2 and output terminal P3 is denoted as... The calculation method is as follows minus According to Kirchhoff's current law, when the bus tie switch is fully closed and the main circuit is continuous, ignoring measurement errors and minimal leakage current, the currents at the three points should theoretically be equal. Therefore... and It should be close to zero. When the switch is open, the current at the internal contact position... The current drops to zero, but there may still be charging current or load current at the input and output terminals, leading to... and Significantly nonzero values ​​appear.

[0056] To quantify the timing consistency of the currents at the three sampling points, a current correlation coefficient is introduced. The correlation coefficient between sampling points P1 and P2 is calculated using a 100-millisecond data window corresponding to the most recent 100 sampling points, employing the Pearson correlation coefficient formula: In the formula, The correlation coefficient between the currents at sampling points P1 and P2 is denoted as , and its value ranges from -1 to 1. The window length for correlation calculation is set to 100. and These are the current values ​​of P1 and P2 at the ks-th sampling time, respectively, in amperes; and , , represent the average currents of P1 and P2 within the window, respectively, in amperes; s is the offset within the window, ranging from 0 to 99; k is the sampling number. When the switch is closed and operating normally, the current changes at the two points show a highly consistent trend. The correlation between the two currents is close to 1; when the switch is open or there is poor contact, the correlation between the two currents weakens. A significant decrease was observed. Similarly, the correlation coefficient between sampling points P2 and P3 was calculated. The correlation coefficient provides time-domain correlation information, which complements the amplitude information of the current difference component, together forming the feature space for state judgment.

[0057] Additionally, the rate of change of current is extracted as a transient feature for rapid identification of switching actions. The rate of change of current at sampling point P2 is calculated as follows: In the formula, Let P2 be the rate of change of current at the k-th sampling time, in amperes per second. and These are the moving average currents of P2 at the k-th and (k-1)-th sampling times, respectively, in amperes; The sampling time interval is in seconds, with a value of 0.001 seconds (1 millisecond); k is the sampling sequence number. The current change rate reflects the dynamic rate of change of the internal contact current. During switch opening and closing, the current change rate will experience a sudden change; during normal operation, the current change rate remains within a stable range. A change rate threshold of 50 amperes per second is set. When the absolute value of the current change rate exceeds this threshold and persists for more than 5 sampling points, it is determined that the switch state has changed, triggering the state confirmation process. These extracted feature quantities include... , , , and These will be used in combination in the state judgment logic of step seven.

[0058] Upon initial power-on or reset, the device automatically executes a system capacity identification procedure. The identification method is based on the float charge current characteristics of the battery pack. With the bus tie switch closed, the input current is monitored. The average value was calculated by continuously collecting data for 60 seconds. In the formula, This is the average float charge current, in amperes. This represents the number of sampling points in the float charge state, with a value of 60000 (60 seconds × 1000 times / second). Let be the moving average current at sampling point P1 at the k-th sampling time, in amperes; k is the sampling number, ranging from 1 to 60000. Based on the empirical relationship between battery float charge current and capacity, the float charge current is approximately one-thousandth to three-thousandths of the battery capacity. The system capacity identification criterion is: when... A capacity less than 0.5 amp-hours is considered a small-capacity system, with a total battery capacity of approximately 200 amp-hours or less; when When the ampere is between 0.5 and 2 amps, it is judged as a medium-capacity system, with a capacity of approximately 200 to 800 amp-hours; when A capacity greater than 2 amp-hours is considered a large-capacity system, with a capacity of 800 amp-hours or more.

[0059] Initialize the current differential threshold based on the identified capacity level. For small-capacity systems, assume 0.1 amps; medium capacity system design It is 0.5 amps; large-capacity system design The threshold is 1.0 amp. This threshold is used to determine whether the differential current component exceeds the normal range. A correlation coefficient threshold is also set according to the capacity level. Small-capacity systems are more sensitive to noise. It is 0.85; for medium and large capacity systems... The value is 0.90. When the measured correlation coefficient is lower than this threshold, the switch state is considered abnormal.

[0060] In addition, the device measures the ambient temperature and obtains the temperature inside the cabinet through the temperature sensor built into the main control unit. The unit is degrees Celsius. Temperature compensation coefficient. Based on temperature calculations, the baseline temperature is set at 25 degrees Celsius. For every 10 degrees Celsius increase or decrease in temperature, the threshold is adjusted by 5%. In the formula, This is the temperature compensation coefficient, which is dimensionless. This is the temperature coefficient, with a value of 0.005 per degree Celsius. This refers to the ambient temperature, expressed in degrees Celsius. The reference temperature is 25 degrees Celsius.

[0061] The temperature-compensated current differential threshold is The correlation coefficient threshold remains unchanged. These initialization parameters are stored in the non-volatile memory of the main control unit and serve as baseline values ​​for adaptive adjustment during subsequent operation, providing initial conditions for the dynamic threshold optimization in step seven.

[0062] During continuous operation, the threshold adaptive unit dynamically optimizes the judgment threshold based on real-time operating data. The optimization strategy is based on statistical learning methods, using historical normal operation data to build a threshold adaptive model. The device records the current difference component in the confirmed closed state of the bus tie switch within the past 24 hours. and Calculate its distribution characteristics and mean. and standard deviation : In the formula, This is the average of the current difference components, in amperes. This represents the standard deviation of the current difference component, in amperes. This is the length of the historical data window, corresponding to the number of sampling points in 24 hours, with a value of 86,400,000 (24 hours × 3,600 seconds / hour × 1,000 times / second). and These represent the current difference component at the h-th sampling time, in amperes; h is the historical data sampling number, ranging from 1 to 86,400,000. Under normal closed-loop conditions, the current difference component should follow a normal distribution with a mean close to zero, and the standard deviation reflects the measurement noise and inherent system fluctuation level.

[0063] Based on the three-standard-deviation principle, the dynamic difference threshold is set to the mean plus three standard deviations: In the formula, This is the adaptive current differential threshold, in amperes. This is the average of the current difference components, in amperes. This represents the standard deviation of the current difference component, in amperes. This threshold covers 99.7% of the normal fluctuation range; current differences exceeding this range are considered abnormal. The dynamic threshold is updated hourly, calculated using a rolling data window of the most recent 24 hours. To prevent abnormal threshold drift, upper and lower limits are set: the dynamic threshold must not exceed twice the initial threshold, nor fall below 0.5 times the initial threshold. If the calculated value exceeds this range, the boundary value is used.

[0064] A similar adaptive strategy is used for the correlation coefficient threshold. This applies to statistical analysis during normal operation. and The distribution of the distribution is calculated, and its mean is determined. and standard deviation Since the ideal correlation coefficient is 1, and normal fluctuations are downward, a one-sided threshold and a dynamic correlation coefficient threshold are used. The threshold is set to the mean minus three standard deviations. Similarly, a constraint is set: the dynamic threshold must not be lower than 0.7 to avoid excessive leniency. Adaptive adjustment enables the device to adapt to changes in characteristics under different operating conditions. For example, increased current fluctuations due to increased battery internal resistance in high summer temperatures, or adjustments in current distribution caused by changes in load characteristics, can all be reasonably tracked through threshold adaptation, improving the robustness and accuracy of detection.

[0065] The state determination unit comprehensively uses the feature quantities extracted in step five and the threshold optimized in this step to execute multi-level judgment logic to determine the actual state of the bus tie switch. The judgment logic is divided into two layers: main criterion and auxiliary criterion. The main criterion is based on the current difference component. When... Less than and Also smaller than When the closing condition of the main criterion is met, it is preliminarily determined that the switch is closed. Greater than or Greater than When the disconnection condition of the main criterion is met, it is initially determined that the switch is open. A threshold of twice the threshold value is introduced as the judgment boundary to provide a judgment margin and avoid oscillation around the threshold.

[0066] The auxiliary criterion is based on the correlation coefficient. When Greater than and Greater than When the closure condition of the auxiliary criterion is met, it indicates that the current change at each point is consistent. Less than or Less than When the disconnection condition of the auxiliary criterion is met, it indicates that the current correlation is lost.

[0067] Comprehensive judgment rules: When the main criterion and the auxiliary criterion simultaneously meet the closing condition, the switch state is confirmed as closed; when the main criterion and the auxiliary criterion simultaneously meet the opening condition, the switch state is confirmed as open; when the conclusions of the main and auxiliary criteria are inconsistent, the system enters the pending confirmation state and extends the observation time to 100 milliseconds. If the criteria are still inconsistent after 100 milliseconds, the main criterion shall prevail, and the inconsistent auxiliary criterion event shall be recorded for fault analysis.

[0068] To avoid false alarms caused by transient interference, a status confirmation delay is implemented. When the judgment logic first detects a status change, it does not immediately output a result but observes for 50 milliseconds, requiring the judgment conclusion to remain consistent during this period. Only when the judgment results for 50 consecutive 50-millisecond samples are all the same status is the status change confirmed as valid, and the output status is updated. This delayed confirmation mechanism effectively filters out the impact of short-term disturbances such as load switching and charging current fluctuations, keeping the false alarm rate within the design specifications. The confirmed switch status is then output to the display alarm unit.

[0069] The alarm display unit uses a TFT touch screen. The status display area shows the current status of the bus tie switch in large font. When closed, it displays "Closed" in green; when open, it displays "Open" in red; and when awaiting confirmation, it displays "Confirming" in yellow.

[0070] The alarm function is divided into three levels: information level, warning level, and fault level. Information level alarms include normal status changes and parameter over-limit prompts, displayed in blue; warning level alarms include insulation degradation, circulating current detection, and yellow health warnings, displayed in yellow and accompanied by audible prompts; fault level alarms include abnormal switch status, severely degraded insulation, and red health warnings, displayed in flashing red and triggering audible and visual alarms.

[0071] Finally, it should be noted that the methods and devices described in detail above are merely embodiments, and those skilled in the art can modify these embodiments in different ways as long as they do not depart from the scope of the present invention.

Claims

1. A smart detection system for the switching state of a DC system based on multi-point current sampling, characterized in that: include, A current sampling unit is used to collect current data of the main circuit of the bus tie switch in the DC system. The current sampling unit includes an array of micro-current sensors deployed at the input terminal P1, internal contact position P2, and output terminal P3 of the bus tie switch, used to detect the real-time current values ​​at both sides and the middle node of the switch, respectively. A signal conditioning and acquisition unit is connected to the current sampling unit and is used to filter and convert the collected current signal into analog and digital signals. A main control processing unit is connected to the signal conditioning and acquisition unit. The main control processing unit is configured to: acquire the current data of each sampling point, calculate the current difference characteristics and current correlation characteristics between adjacent sampling points, and combine them with preset state judgment logic to determine the actual on or off state of the bus tie switch.

2. The intelligent detection system for DC system switching status based on multi-point current sampling according to claim 1, characterized in that: The sensor at the input terminal P1 is located at the connection point connecting the first battery pack to the input terminal of the bus tie switch; the sensor at the output terminal P3 is located at the connection point connecting the output terminal of the bus tie switch to the second battery pack; and the sensor at the internal contact location P2 is located near the internal contact of the switch. These three sampling points are connected in series in spatial topology, and the sensors in the microcurrent sensor array are Hall current sensors.

3. The intelligent detection system for DC system switching status based on multi-point current sampling according to claim 2, characterized in that: It also includes a threshold adaptive unit, which is used to dynamically adjust the judgment threshold in the state judgment logic; when the device is started, the capacity level of the DC system is identified according to the average value of the monitored floating charge current, and the current difference threshold and correlation coefficient threshold are initialized according to the capacity level. During device operation, the distribution characteristics of current differential components under historical normal closed states are statistically analyzed, their mean and standard deviation are calculated, and the dynamic current differential threshold is updated in real time.

4. The intelligent detection system for DC system switching status based on multi-point current sampling according to claim 3, characterized in that: When extracting features, the main control processing unit needs to calculate the first current difference component Δ between the input terminal P1 and the internal contact position P2. I 12 And the second current difference component Δ between the internal contact position P2 and the output terminal P3. I 23 ; Calculate the first Pearson correlation coefficient r between input terminal P1 and internal contact position P2 within a set time window. 12 And the second Pearson correlation coefficient r between the internal contact position P2 and the output terminal P3. 23； The state judgment logic is as follows: when the current difference component is less than the dynamic current difference threshold and the correlation coefficient is greater than the correlation coefficient value, the switch is determined to be closed. When the current difference component is greater than the dynamic current difference threshold or the correlation coefficient is less than the correlation coefficient value, the switch is determined to be open.

5. The intelligent detection system for DC system switching status based on multi-point current sampling according to claim 4, characterized in that: The threshold adaptive unit also includes a temperature compensation module; the device is equipped with a temperature sensor, and the temperature compensation module corrects the current differential threshold using a temperature coefficient based on the collected ambient temperature, with the reference temperature set to 25 degrees Celsius.

6. The intelligent detection system for DC system switching status based on multi-point current sampling according to claim 5, characterized in that: It also includes an insulation monitoring unit and a transfer switch unit; the current sampling unit further includes feeder current sensors F1 to Fn installed at the outgoing ends of each DC feeder; the insulation monitoring unit uses the AC injection method to inject a low-frequency AC signal into the DC bus, and uses the feeder current sensors to detect the AC response current of each feeder to calculate the insulation resistance to ground; the transfer switch unit includes a main bus switch group and a backup bus switch group; when the main control processing unit receives a signal that the insulation resistance of a feeder is lower than the alarm threshold, it controls the transfer switch unit to seamlessly switch the faulty feeder from the main bus to the backup bus.

7. The intelligent detection system for DC system switching state based on multi-point current sampling according to claim 6, characterized in that: The main control processing unit is also configured to perform DC loop detection function, collect the output current of the two sets of batteries and the load current of all feeders; calculate the difference between the total output current of the batteries and the total load current of the feeders; when the absolute value of the difference continuously exceeds the circulating current judgment threshold, it is determined that there is circulating current in the system, and the circulating current path is located by combining the current direction information of the internal contact position P2 of the bus tie switch.

8. The intelligent detection system for DC system switching state based on multi-point current sampling according to claim 7, characterized in that: The main control processing unit is also equipped with a status confirmation delay mechanism: when a change in the switch status is detected, it continuously observes for a preset time length, and only when the judgment results of all sampling points are consistent within this time period is the switch status output confirmed and updated to filter out transient interference.

9. The intelligent detection system for DC system switching state based on multi-point current sampling according to claim 8, characterized in that: The signal conditioning and acquisition unit performs first-order low-pass filtering on each channel signal, with the filter cutoff frequency set to 100Hz to attenuate high-frequency noise interference; the sampling frequency of the current sampling unit is set to 1kHz.

10. The intelligent detection system for DC system switching state based on multi-point current sampling according to claim 9, characterized in that: It also includes a display warning unit, which uses a TFT touch screen to indicate the switch's closed, open, and pending confirmation states with different colors. The alarm function is divided into three levels: information level, warning level, and fault level, each corresponding to different display colors and sound and light prompt strategies.

Images