A secondary water supply operation data processing method and system based on multi-dimensional analysis

Abstract

The application discloses a kind of secondary water supply operation data processing method and system based on multidimensional analysis, it is related to secondary water supply technical field, this method is by constructing frequency excitation source perception factor Afri, combined with current disturbance acceleration AIi, pressure echo delay time Tpr and frequency matching characteristic factor, traditional resonance result is identified in advance to resonance excitation trend feedforward prediction, avoid the problem that traditional dependence mechanical sensor, such as vibration instrument, response lag and false alarm rate is high.This method is based on the operation data itself and constructs pure data-driven perception mechanism, can issue early warning before resonance has not yet formed but frequency behavior enters high excitation interval, and through the upper limit value Afri max And excitation grade division mechanism of excitation source perception factor, complete the continuous monitoring and graded identification of resonance state, significantly enhance the real-time perception ability and risk response ability of water pump operation state.

Description

Technical Field

[0001] This invention relates to the field of secondary water supply technology, specifically to a method and system for processing secondary water supply operation data based on multi-dimensional analysis. Background Technology

[0002] Secondary water supply operation data processing involves data processing technology in intelligent pump operation control systems, specifically applied to the identification and dynamic adjustment of frequency behavior in the parallel operation of multiple pumps in secondary water supply systems. Secondary water supply, a common method in urban buildings, commercial areas, residential buildings, and high-rise buildings, relies heavily on pump units under variable frequency control to provide continuous and stable water pressure. Especially in multi-pump parallel water supply modes, the frequency variation paths of different pumps can easily create dynamic interference in the system. The energy coupling behind frequency disturbances is often the direct source of water hammer, oscillations, and acoustic resonance, affecting system operating efficiency and stability. Therefore, identifying, assessing, and proactively avoiding potential resonance excitation frequency ranges through real-time operational data is crucial for improving the resilience and intelligence level of intelligent water supply systems.

[0003] Currently, in existing technologies, secondary water supply systems mostly rely on fixed adjustment parameters or simple threshold methods for frequency control. This fails to detect subtle disturbances in pump operation or identify triggering frequency points in real time based on the system's current dynamic state, resulting in control disadvantages such as "blind frequency adjustment" or "response lag." More critically, current systems often use mechanical vibration or noise sensing devices, such as vibration sensors, to detect resonance. These methods are complex to deploy, prone to false alarms, and costly. Especially under low-load operating conditions at night, when the system enters a "low-frequency resonance sensitive state," multiple pump frequencies may repeatedly fluctuate within a small range, forming a so-called "frequency oscillation loop," which can, in severe cases, induce continuous water hammer effects.

[0004] The root cause of the above problems lies in the fact that frequency disturbances in secondary water supply systems are caused by the superposition of multiple weak factors, such as nonlinear changes in current, unstable instantaneous pressure feedback, and frequency excitation matching resonance frequency ranges. These factors are highly concealed and cannot be accurately diagnosed by relying solely on a single sensor signal. When the system operating frequency gradually approaches the potential resonance frequency range, if it is not identified and adjusted in advance, it may trigger a strong system response, such as severe pump vibration, current jumps, and pressure wave rebounds. Ultimately, this leads to severe system frequency fluctuations, flow instability, water pressure fluctuations, and a surge in noise. Especially during off-peak water usage periods at night, low-frequency coupling resonance of multiple pumps occurs frequently, causing the frequency converter control system to fall into a high-frequency frequency hopping state, seriously affecting the system's energy-saving effect and service life, and even triggering safety risks such as false alarm shutdowns and equipment damage. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method and system for processing secondary water supply operation data based on multi-dimensional analysis, which solves the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution, comprising the following steps:

[0007] S1. By embedding an API application interface in the central controller of the secondary water supply pump, the operating data is collected in real time and transmitted to the operating data processing server through the API application interface.

[0008] S2. In the running data processing server, preprocess the running data to obtain a standardized dataset, and perform calculations based on the standardized dataset to output the excitation source perception factor Afri.

[0009] S3. Extract the upper limit value of the excitation source sensing factor Afri based on the excitation source sensing factor Afri. max And set the threshold for the resonance excitation range to conduct a preliminary comparative evaluation;

[0010] S4. Based on the preliminary comparison and evaluation results, trigger the avoidance frequency window construction mechanism and construct the avoidance frequency window function Wavoid.

[0011] S5. Based on the excitation source sensing factor Afri and the avoidance frequency window function Wavoroid of each water pump, perform summary calculations, output the adjusted target frequency Fadj, and execute rhythm control and anti-oscillation protection mechanisms.

[0012] Preferably, S1 includes S11 and S12;

[0013] S11. By embedding an API application interface in the pump center controller of the secondary water supply and deploying an API client in the pump center controller through the API application interface, the pump frequency conversion control terminal, pump motor terminal and resonance response recording module of the pump center controller are automatically connected.

[0014] After the connection is completed, the API client will automatically run to send a data acquisition command to the water pump central controller to collect the water pump's operating data in the secondary water supply process in real time.

[0015] The API application interface establishes connections with the PLC control system of the water pump central controller, the ABB water pump smart meter, and the water pump pressure valve.

[0016] The operational data includes the output frequency f of the i-th water pump at time t. i (t) The current disturbance acceleration AI of the i-th water pump at time t i(t) and the pressure echo delay time Tpr(t) at time t;

[0017] S12. Configure the network access function of the API application interface, use the HTTPS transmission protocol to establish a network connection between the running data processing server and the API application interface, and transmit the running data to the running data processing server over the network.

[0018] Preferably, S2 includes S21;

[0019] S21. Preprocess the running data in the running data processing server to obtain a standardized dataset;

[0020] The preprocessing includes time series standardization, data cleaning, and normalization.

[0021] The time series standardization uses a regular sampling rearrangement method to uniformly timestamp the sampling time points of all parameters in the running data;

[0022] The data cleaning process involves filling in missing values ​​using a sliding window mean when the continuous missing time in the running data is ≤3 seconds, and directly removing the corresponding values ​​if the missing time exceeds 10 seconds.

[0023] Preferably, S2 further includes S22;

[0024] S22. Start the water pump for secondary water supply. As the water pump's current secondary water supply status changes, continuously update the standardized dataset. At the same time, build a frequency excitation source sensing algorithm model for the water pump in the running data processing server. Input the real-time updated standardized dataset into the frequency excitation source sensing algorithm model for calculation and output the excitation source sensing factor Afri for each water pump to measure the degree of risk of resonance induction by the operating frequency of the water pump during the secondary water supply process.

[0025] The excitation source sensing factor Afri is calculated and output using the following frequency excitation source sensing algorithm model;

[0026] ;

[0027] In the formula, Afri i (t) represents the excitation source sensing factor of the i-th water pump at time t, sin represents the cosine function, π represents pi and takes the value 3.14, and fres represents the center value of the frequency cluster fitting of the water pump resonance event and takes the value without dimension.

[0028] Preferably, S3 includes S31;

[0029] S31. Based on the excitation source sensing factor Afri of each water pump, extract the maximum value within a 60-second sliding window to obtain the upper limit value Afri of the excitation source sensing factor.max The upper limit of the excitation source sensing factor Afri max The specific extraction formula is as follows: In the formula, Afri max (i) represents the upper limit of the sensing factor of the excitation source of the i-th water pump, max means taking the upper limit value, and T represents the sliding window time length.

[0030] Preferably, S3 includes S32;

[0031] S32. Extract historically confirmed resonance time samples, analyze the distribution of the excitation source sensing factor Afri in the resonance time samples, and extract the resonance excitation interval threshold, which includes a first resonance excitation threshold F1 and a second resonance excitation threshold F2. Then, use the real-time acquired upper limit value Afri of the excitation source sensing factor of the i-th water pump. max (i) A preliminary comparison and evaluation is conducted with the resonance excitation interval threshold, and the resonance excitation level is classified based on the preliminary comparison and evaluation results to determine the resonance sensitivity state of the current water pump during secondary water supply. Then, based on the resonance excitation level classification results, an avoidance frequency window construction mechanism is triggered. The specific evaluation content is as follows:

[0032] When the upper limit of the sensing factor of the excitation source of the i-th water pump is Afri max (i) When the frequency is less than the first resonance excitation threshold F1, it indicates a safe frequency band. At this time, it is classified as Level I and no intervention is required for normal operation.

[0033] When the first resonance excitation threshold F1 ≤ the upper limit value of the sensing factor of the i-th water pump excitation source Afri max (i) When < the second resonance excitation threshold F2, it indicates the secondary excitation frequency band, which is classified as Level II. At this time, the sampling frequency of the API application interface is increased by 50%.

[0034] When the upper limit of the sensing factor of the excitation source of the i-th water pump is Afri max (i) When the second resonance excitation threshold F2 is greater than or equal to the second resonance excitation threshold, it indicates the resonance excitation frequency band. At this time, it is classified as Level III, triggering the avoidance frequency window construction mechanism.

[0035] 7. Preferably, S4 includes S41 and S42;

[0036] S41. After a preliminary comparative evaluation of the trigger avoidance frequency window construction mechanism, the output frequency f of the i-th water pump at time t is used. i (t) Perform discrete binning processing to divide the entire frequency range into several non-overlapping intervals to obtain frequency segment x;

[0037] Based on the preliminary comparative evaluation of the resonance excitation level classification results, for each frequency band x, the upper limit value of the sensing factor Afri of the excitation source of the i-th water pump is used. max(i) The resonance excitation level classification results are labeled with Level I, Level II and Level III, and the frequency band status label Lf(x) of frequency band x is obtained.

[0038] Simultaneously, the upper limit value of the sensing factor for the excitation source of the i-th water pump, Afri, was selected. max (i) The time and corresponding frequency of the second resonance excitation threshold F2 are mapped into the frequency segment x. The number of frequencies belonging to the resonance excitation frequency segment on the frequency segment x is counted. The perturbation density Df(x) of the frequency segment x is obtained by dividing by the sliding window time length T.

[0039] S42. Based on the frequency band status label Lf(x) and the perturbation density Df(x) of frequency band x, construct the avoidance frequency window function Wavoid, and determine the resonance to avoid frequency band x.

[0040] The specific construction method of the avoidance frequency window function Wavoroid is as follows:

[0041] ;

[0042] In the formula, Wavoid(x) represents the avoidance frequency window function for frequency band x, and otherwise means the opposite. The avoidance frequency window function Wavoid=1 means that avoidance is required, and the avoidance frequency window function Wavoid=0 means that no avoidance is required and free operation is allowed. The perturbation density Df(x) of frequency band x > 95% means that the perturbation density Df(x) of frequency band x exceeds 95% of the excitation density, and is judged as a density anomaly.

[0043] Preferably, S5 includes S51;

[0044] S51. Calculate the adjusted target frequency Fadj based on the current avoidance frequency window function Wavoid and the excitation source sensing factor Afri of each water pump, and then transmit the adjusted target frequency Fadj to the frequency converter to adjust the current water pump operating frequency.

[0045] The target frequency Fadj is calculated and output using the following algorithm formula;

[0046] ;

[0047] In the formula, Fadj i (t) represents the target frequency of the i-th water pump at time t, and d represents the integral function. This represents the adjustment response factor, used to control the adjustment speed. It is configured by the user and has a dimensionless value, Wavoid(f). i (t) represents the output frequency f of the i-th water pump at time t. i The avoidance frequency window function of (t).

[0048] Preferably, S5 further includes S52;

[0049] S52. After adjusting the pump operating frequency through the target frequency Fadj, a rhythm control and anti-oscillation protection mechanism is executed. The rhythm control and anti-oscillation protection mechanism includes frequency locking, minimum adjustment interval, avoidance trigger conditions, and convergence judgment.

[0050] The frequency locking is achieved by locking the adjusted target frequency Fadj, with the adjustment range limited to ±1%.

[0051] The minimum adjustment interval is defined by limiting the interval between each adjustment operation to at least 10 seconds;

[0052] The avoidance trigger condition is that if the frequency jump is still at level III after three consecutive frequency jumps, the forced avoidance mechanism will be activated, directly limiting the adjustment range to within 5% above and below, and quickly leaving the risk area.

[0053] The convergence judgment is made by determining that the adjustment strategy is effective if the frequency is below Level II for three consecutive cycles, adding the current frequency to the safe operation frequency band list, turning off the adjustment, and restoring the water pump to normal operation.

[0054] A secondary water supply operation data processing system based on multi-dimensional analysis includes an operation data acquisition module, a frequency excitation source sensing module, a resonance excitation level classification module, an avoidance frequency construction module, and a dynamic frequency adjustment module.

[0055] The operation data acquisition module acquires operation data in real time by embedding an API application interface in the central controller of the secondary water supply pump, and transmits the operation data to the operation data processing server through the API application interface.

[0056] The frequency excitation source sensing module preprocesses the running data in the running data processing server to obtain a standardized dataset, and calculates based on the standardized dataset to output the excitation source sensing factor Afri.

[0057] The resonance excitation level classification module extracts the upper limit value of the excitation source sensing factor Afri based on the excitation source sensing factor Afri. max And set the threshold for the resonance excitation range to conduct a preliminary comparative evaluation;

[0058] The avoidance frequency construction module triggers the avoidance frequency window construction mechanism based on the preliminary comparison and evaluation results, and constructs the avoidance frequency window function Wavoid.

[0059] The dynamic frequency adjustment module performs a summary calculation based on the excitation source sensing factor Afri and the avoidance frequency window function Wavoroid of each water pump, outputs the adjusted target frequency Fadj, and executes rhythm control and anti-oscillation protection mechanisms.

[0060] This invention provides a method and system for processing secondary water supply operation data based on multi-dimensional analysis. It has the following beneficial effects:

[0061] (1) This method constructs a frequency excitation source sensing factor Afri, combined with current disturbance acceleration AIi, pressure echo delay time Tpr, and frequency matching characteristic factors, shifting the traditional reliance on resonance result identification to feedforward prediction of resonance excitation trend. This avoids the problems of delayed response and high false alarm rate associated with traditional reliance on mechanical sensors, such as vibrators. This method constructs a purely data-driven sensing mechanism based on the operational data itself, enabling early warning before resonance forms but before the frequency behavior enters the high excitation range. The upper limit value of the excitation source sensing factor Afri is used to determine the warning. max It also triggers a classification mechanism to achieve continuous monitoring and hierarchical identification of resonance states, significantly enhancing the real-time perception and risk response capabilities of water pump operation status.

[0062] (2) This method constructs a resonant segment identification and shielding mechanism based on density and level by jointly evaluating the perturbation density Df(x) and the frequency band status label Lf(x) of each frequency band x. This mechanism can effectively filter out pseudo-high-frequency perturbation regions in the frequency band and only implement the avoidance strategy for truly dangerous frequency bands with high density and high level, thus avoiding the adjustment oscillation caused by blind frequency jumps. Furthermore, a gradient adjustment model is constructed through the target frequency Fadj, and the optimal adjustment direction is calculated based on the differential characteristics of the excitation source sensing factor Afri and the frequency, so that the frequency adjustment is transformed from a static strategy to a dynamic response path, forming an adjustment closed loop of avoiding frequency band shielding and gradually approaching the safe frequency zone, which effectively suppresses the random fluctuations of the system frequency.

[0063] (3) This method constructs a closed-loop control logic for the entire process, including frequency locking interval, minimum adjustment interval, resonance trigger defense, and adjustment strategy convergence judgment, through rhythmic control and anti-oscillation protection mechanism driven by the adjusted target frequency Fadj. Especially when the system enters the low-load state of nighttime resonance, it can effectively avoid secondary fluctuation problems caused by drastic frequency jumps. By automatically adjusting the adjustment step size, widening the avoidance frequency band, and automatically sealing the current frequency as a safe operating zone after the strategy is effective, the adjustment behavior is more rhythmic and stable, significantly reducing the problems of false frequency jumps, frequent alarms, and load fluctuations in the system, and improving the overall intelligence level and stable operation capability of the system. Attached Figure Description

[0064] Figure 1 This is a schematic diagram of the steps of a secondary water supply operation data processing method based on multi-dimensional analysis according to the present invention;

[0065] Figure 2 This is a schematic diagram of a secondary water supply operation data processing system based on multi-dimensional analysis according to the present invention.

[0066] Figure 3 This is a schematic diagram showing the composition and data flow path of a secondary water supply operation data processing server. Detailed Implementation

[0067] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0068] Example 1

[0069] Please see Figure 1 and Figure 3 This invention provides a method for processing secondary water supply operation data based on multi-dimensional analysis. To achieve the above objectives, this invention is implemented through the following technical solution, including the following steps:

[0070] S1. By embedding an API application interface in the central controller of the secondary water supply pump, the operating data is collected in real time and transmitted to the operating data processing server through the API application interface.

[0071] S2. In the running data processing server, preprocess the running data to obtain a standardized dataset, and perform calculations based on the standardized dataset to output the excitation source perception factor Afri.

[0072] S3. Extract the upper limit value of the excitation source sensing factor Afri based on the excitation source sensing factor Afri. max And set the threshold for the resonance excitation range to conduct a preliminary comparative evaluation;

[0073] S4. Based on the preliminary comparison and evaluation results, trigger the avoidance frequency window construction mechanism and construct the avoidance frequency window function Wavoid.

[0074] S5. Based on the excitation source sensing factor Afri and the avoidance frequency window function Wavoroid of each water pump, perform summary calculations, output the adjusted target frequency Fadj, and execute rhythm control and anti-oscillation protection mechanisms.

[0075] In this embodiment, the method embeds an API application interface into the central controller of the pump in the secondary water supply system to construct a real-time acquisition mechanism covering multiple types of unconventional operating parameters, such as output frequency, dynamic current disturbance, and pressure response delay. This operational data is then efficiently transmitted to an operational data processing server for centralized analysis and modeling. On the data processing server, through standardization and outlier cleaning, a dynamically updated dataset is constructed and input into the frequency excitation source sensing algorithm model. This model outputs an induced intensity index between the pump operating frequency and system resonance, namely the excitation source sensing factor Afri. A sliding window is then used to extract its upper limit value, Afrimax, and combined with historical sample thresholds F1 and F2 for grading, achieving quantitative identification and preliminary classification of potential resonance risks. Furthermore, the method further extracts high-risk frequency bands for resonance excitation by constructing frequency band disturbance density Df and frequency band state labels Lf, generating a avoidance frequency window function, Wavoroid, to achieve precise avoidance and isolation of high-risk frequency bands. Finally, based on the changing trend of the current excitation source sensing factor Afri and the avoidance frequency window function Wavorid shielding mechanism, a target frequency Fadj is jointly generated to drive the dynamic adaptive adjustment of the pump's operating frequency. This is accompanied by the setting of a frequency locking range, minimum adjustment interval, and convergence judgment mechanism to ensure the controllable rhythm of the adjustment path and prevent frequency hopping oscillations and resonance re-triggering. Through the above implementation methods, this approach not only achieves accurate identification and predictive control of pump operating resonance risks without relying on traditional vibration monitoring hardware, but also significantly improves the diagnostic speed of resonance response, the intelligence of frequency adjustment, and the stability of strategy control. This significantly enhances the overall safety, reliability, and adaptive operation capability of the system under low-load nighttime operation scenarios, providing a forward-looking and robust technical solution for water supply scheduling under complex operating conditions.

[0076] Example 2

[0077] Please see Figure 1 and Figure 3 Specifically: S1 includes S11 and S12;

[0078] S11. By embedding an API application interface in the pump center controller of the secondary water supply and deploying an API client in the pump center controller through the API application interface, the pump frequency conversion control terminal, pump motor terminal and resonance response recording module of the pump center controller are automatically connected.

[0079] After the connection is completed, the API client will automatically run to send a data acquisition command to the water pump central controller to collect the water pump's operating data in the secondary water supply process in real time.

[0080] The API application interface establishes connections with the PLC control system of the pump center controller, ABB pump smart meters, and pump pressure valves.

[0081] The operational data includes the output frequency f of the i-th water pump at time t. i (t) The current disturbance acceleration AI of the i-th water pump at time t i (t) and the pressure echo delay time Tpr(t) at time t;

[0082] The output frequency f of the i-th water pump at time t i (t) Establish a connection with the PLC control system of the secondary water supply pump frequency converter through the embedded API application interface, and extract the frequency values ​​of the output of multiple pumps within a time period.

[0083] The current disturbance acceleration AI of the i-th water pump at time t i (t) A connection is established with the ABB smart meter at the motor end of the secondary water supply pump via the embedded API application interface, the original current value of each pump is extracted, and the second derivative is calculated to obtain the value.

[0084] The pressure echo delay time Tpr(t) at time t is connected to the resonance response recording module of the water pump pressure valve through the embedded API application interface. The pressure disturbance of the water pump at the front and back ends after the water pump is excited to resonance is extracted in real time, and the propagation time of the pressure disturbance between the front and back ends is extracted.

[0085] S12. Configure the network access function of the API application interface, use the HTTPS transmission protocol to establish a network connection between the running data processing server and the API application interface, and transmit the running data to the running data processing server over the network.

[0086] In this embodiment, the method embeds an API application interface into the central controller of the secondary water supply system's pumps and deploys an API client, achieving automatic access and seamless integration with the pump frequency converter control terminal, pump motor terminal, and resonance response recording module. This constructs a scalable, high-concurrency, and loosely coupled operational data acquisition architecture. Specifically, the API application interface establishes connections with the PLC control system, the ABB pump smart meter, and the resonance response recording module of the pump pressure valve, respectively. It collects the output frequency f, current disturbance acceleration AI, and pressure echo delay time Tpr of each pump in real time, and performs deep feature extraction on the raw signals using the second derivative and pressure disturbance propagation algorithm. After acquisition, the data is stably and securely transmitted to the operational data processing server via the configured HTTPS network transmission protocol of the API, forming a unified data processing entry point. This method significantly reduces system instability and maintenance costs caused by independent data acquisition from multiple sensors, complex communication, and incompatible interfaces in traditional secondary water supply systems. Furthermore, through an integrated API acquisition mechanism, it unifies the previously scattered dynamic response data in terms of timing and structure, allowing it to be directly used in subsequent multi-dimensional feature analysis, resonance identification, and frequency optimization control strategies. This not only significantly improves the real-time performance, accuracy, and continuity of data acquisition but also provides a solid data foundation for building high-dimensional analysis models and intelligent control mechanisms, thereby enhancing the system's intelligence level, operational safety, and resilience against complex disturbances under low nighttime loads.

[0087] Example 3

[0088] Please see Figure 1 and Figure 3 Specifically: S2 includes S21;

[0089] S21. Preprocess the running data in the running data processing server to obtain a standardized dataset;

[0090] Preprocessing includes time-series standardization, data cleaning, and normalization.

[0091] Time series standardization uses regular sampling rearrangement to assign a unified timestamp to the sampling time points of all parameters in the running data;

[0092] Data cleaning involves filling missing values ​​with a sliding window mean when the continuous missing time in the running data is ≤3 seconds, and directly removing the corresponding values ​​if the missing time exceeds 10 seconds to prevent misjudgment.

[0093] S2 also includes S22;

[0094] S22. Start the water pump for secondary water supply. As the water pump's current secondary water supply status changes, continuously update the standardized dataset. At the same time, build a frequency excitation source sensing algorithm model for the water pump in the running data processing server. Input the real-time updated standardized dataset into the frequency excitation source sensing algorithm model for calculation and output the excitation source sensing factor Afri for each water pump to measure the degree of risk of resonance induction by the operating frequency of the water pump during the secondary water supply process.

[0095] The excitation source sensing factor Afri is calculated and output using the following frequency excitation source sensing algorithm model;

[0096] ;

[0097] In the formula, Afri i (t) represents the excitation source sensing factor of the i-th water pump at time t, sin represents the cosine function, π represents pi, with a value of 3.14, and fres represents the center value of the frequency cluster fitting of the water pump resonance event, with a dimensionless value.

[0098] The derivation logic of the formula: the current disturbance acceleration AI of the i-th water pump at time t. i (t) is used to reflect whether the trend of current change is rapidly accelerating or decelerating. A large current disturbance acceleration indicates that the water pump load state has an acceleration level fluctuation, which is usually caused by resonance or sudden change.

[0099] This represents the reciprocal of the pressure echo delay time. The smaller the delay, the faster the response. The larger this value, the faster the response. A rapid response indicates that the system resonance point has been triggered, which is highly dangerous.

[0100] This represents the resonance matching factor, which measures the degree of matching between the current frequency and the easily excitable resonant frequency. When this value is 0, it indicates the strongest matching.

[0101] The derivation of the formula reveals three core causes of resonance: the frequency being within a "dangerous bandwidth"; a violent response to system load; and rapid and transient anomalies in pressure feedback. These three core causes can be abstracted as the current disturbance acceleration AI of the i-th pump at time t. i (t), the reciprocal of the pressure echo delay time and the resonance matching factor, the three are multiplied together to form the resonance excitation sensing factor;

[0102] The physical meaning of the formula: In a secondary water supply system, multiple pumps connected in parallel may experience resonance-induced problems when operating at low frequency or under low load at night, such as pressure wave jitter, current jump, and water hammer delay. A frequency excitation source sensing algorithm model is constructed to calculate the excitation source sensing factor Afri, which is used to identify in real time whether a certain frequency point may excite system resonance, thereby avoiding that frequency range in advance and ensuring stable system operation.

[0103] In this embodiment, the method uniformly rearranges various data sampling points through time-series standardization, ensuring alignment of multi-source asynchronous data in the time dimension. A combined data cleaning strategy of sliding window mean filling and anomaly removal effectively reduces the risk of misjudgment caused by short-term packet loss and communication delays, improving overall data stability and reliability. Based on this standardized dataset, the running data processing server further embeds a frequency excitation source sensing algorithm model to calculate the excitation source sensing factor Afri for each pump in real time, forming a dynamic quantitative evaluation mechanism for whether the current operating frequency poses a risk of resonance. This implementation not only transforms raw, messy data into high-quality structured data but also provides crucial sensing basis for subsequent resonance judgment, frequency adjustment, and avoidance through the dynamic output of the excitation source sensing factor Afri. Its technical effects are as follows: it significantly improves the system's ability to proactively identify potential resonance hazards during low-load and non-operating nighttime periods, enhancing the intelligence and response accuracy of system regulation; at the same time, the excitation source sensing factor Afri serves as a dynamic safety assessment indicator, providing quantitative feedback support for multi-pump collaborative frequency optimization and anti-oscillation strategy execution, fundamentally improving the stability, robustness, and operating efficiency of the entire secondary water supply system.

[0104] Example 4

[0105] Please see Figure 1 Specifically: S3 includes S31 and S32;

[0106] S31. Based on the excitation source sensing factor Afri of each water pump, extract the maximum value within a 60-second sliding window to obtain the upper limit value Afri of the excitation source sensing factor. max The upper limit of the source sensing factor Afri max The specific extraction formula is as follows: In the formula, Afri max (i) represents the upper limit of the sensing factor of the excitation source of the i-th water pump, max means taking the upper limit value, and T represents the sliding window time length, which is 60 seconds here;

[0107] The reason for taking the maximum value is that resonance is usually caused by short-duration strong excitation, and the maximum value can capture the sudden high excitation frequency point.

[0108] S3 includes S32;

[0109] S32. Extract historically confirmed resonance time samples, analyze the distribution of the excitation source sensing factor Afri in the resonance time samples, extract the resonance excitation interval threshold, which includes the first resonance excitation threshold F1 and the second resonance excitation threshold F2, and then use the real-time acquired upper limit value Afri of the excitation source sensing factor of the i-th water pump. max (i) A preliminary comparison and evaluation is conducted with the resonance excitation interval threshold, and the resonance excitation level is classified based on the preliminary comparison and evaluation results to determine the resonance sensitivity state of the current water pump during secondary water supply. Then, based on the resonance excitation level classification results, an avoidance frequency window construction mechanism is triggered. The specific evaluation content is as follows:

[0110] The first resonance excitation threshold F1 is the upper limit of the stable operating region of historically confirmed resonance time samples;

[0111] The second resonance excitation threshold F2 is passed through the lower limit of the concentrated area of ​​resonance alarm events;

[0112] When the upper limit of the sensing factor of the excitation source of the i-th water pump is Afri max (i) When < the first resonance excitation threshold F1, it indicates a safe frequency band. At this time, it is classified as Level I. Normal operation does not require intervention. That is, the current operating frequency will not cause amplification resonance behavior. No avoidance or adjustment is required. It belongs to the default adjustment frequency band preferred area.

[0113] When the first resonance excitation threshold F1 ≤ the upper limit value of the sensing factor of the i-th water pump excitation source Afri max (i) When < the second resonance excitation threshold F2, it indicates the secondary excitation frequency band. At this time, it is classified as Level II. At this time, the sampling frequency of the API application interface is increased by 50%. The current change has a certain trend, but no sudden change occurs. The delay decreases, indicating that the water pump response is accelerating. The current frequency may be close to the resonance point but has not been triggered.

[0114] When the upper limit of the sensing factor of the excitation source of the i-th water pump is Afri max (i) When the second resonance excitation threshold F2 is greater than or equal to the second resonance excitation threshold, it indicates the resonance excitation frequency band. At this time, it is classified as Level III, triggering the avoidance frequency window construction mechanism.

[0115] In this embodiment, the method sets the sliding window length to 60 seconds, extracts the maximum value of the excitation source sensing factor Afri for each water pump within that window, and forms the upper limit value of the excitation source sensing factor Afri. maxThis allows for the precise capture of short-term, strong excitation behaviors, avoiding the masking of sudden resonance risks by using the mean. Subsequently, by analyzing a large number of historical resonance event samples, a first resonance excitation threshold F1 and a second resonance excitation threshold F2 are defined, thereby constructing a static, empirical safety boundary to achieve a graded judgment of the current operating state. This mechanism not only classifies the current frequency operating state into safe frequency band I, secondary excitation frequency band II, and resonant excitation frequency band III, but also triggers subsequent frequency avoidance strategies accordingly. The beneficial effects of this implementation are reflected in three aspects: first, it improves the system's sensitivity to sudden excitation signals through the sliding window maximum value extraction method; second, it enhances the system's adaptive risk identification capability by constructing a threshold system based on historical samples; and third, the introduction of the graded classification mechanism enables quantitative, hierarchical management of pump operating states, providing a scientific and clear triggering basis for subsequent avoidance mechanisms and frequency adjustment strategies. Ultimately, this improves the operational safety, stability, and intelligent response level of the secondary water supply system under complex and fluctuating operating conditions.

[0116] Example 5

[0117] Please see Figure 1 Specifically: S4 includes S41 and S42;

[0118] S41. After a preliminary comparative evaluation of the trigger avoidance frequency window construction mechanism, the output frequency f of the i-th water pump at time t is used. i (t) Perform discrete binning processing to divide the entire frequency range into several non-overlapping intervals and obtain frequency segments x. Use these segments to calculate characteristic indicators for each frequency x individually to determine whether it belongs to a certain region, such as 38.0~38.5Hz, 38.5~39.0Hz.

[0119] Based on the preliminary comparative evaluation of the resonance excitation level classification results, for each frequency band x, the upper limit value of the sensing factor Afri of the excitation source of the i-th water pump is used. max (i) The resonance excitation level classification results are labeled with Level I, Level II and Level III, and the frequency band status label Lf(x) of frequency band x is obtained; the significance of introducing the frequency band status label Lf(x) of frequency band x is that relying solely on the number of frequency band excitations may result in pseudo-high frequency areas, such as multiple slight disturbances. Only those frequency bands with dense excitation and high level constitute a real risk. This is the key mechanism for filtering interference frequency bands and ensuring the stability of the strategy.

[0120] Simultaneously, the upper limit value of the sensing factor for the excitation source of the i-th water pump, Afri, was selected. max(i) The time and corresponding frequency of the second resonance excitation threshold F2 are mapped to frequency segment x. The number of frequencies belonging to the resonance excitation frequency band in frequency segment x is counted and divided by the sliding window time length T to obtain the perturbation density Df(x) of frequency segment x. The significance of introducing the perturbation density Df(x) of frequency segment x is to represent the high incidence rate of excitation frequency in frequency segment x. The higher the value, the more frequently the water pump enters the potential resonance state in this frequency segment x. It is a data-driven resonance frequency identification method. The significance of introducing the perturbation density Df(x) of frequency segment x is that it is related to the upper limit value of the excitation source sensing factor Afri of the i-th water pump. max (i) The instantaneous value cannot reflect the trend. The density dimension is added to find those persistent resonance sensitive areas, rather than accidental excitation points, and to improve resonance identification from single-point behavior to frequency band x behavior statistics.

[0121] S42. Based on the frequency band status label Lf(x) and the perturbation density Df(x) of frequency band x, construct the avoidance frequency window function Wavoid, and determine the resonance to avoid frequency band x.

[0122] The specific construction method of the avoidance frequency window function Wavoid is as follows:

[0123] ;

[0124] In the formula, Wavoid(x) represents the avoidance frequency window function for frequency band x, and otherwise means the opposite. The avoidance frequency window function Wavoid=1 means that avoidance is required, and the avoidance frequency window function Wavoid=0 means that no avoidance is required and free operation is allowed. The perturbation density Df(x) of frequency band x > 95% means that the perturbation density Df(x) of frequency band x exceeds 95% of the excitation density, and is judged as a density anomaly.

[0125] In this embodiment, the method uses the output frequency f of the i-th water pump at time t. i (t) Discrete binning is performed, dividing the frequency into multiple fixed-width frequency bands x. Combined with the previous resonant excitation level classification results, each frequency band x is assigned a frequency band state label Lf(x). This labeling mechanism effectively filters out pseudo-excitation regions caused by occasional disturbances, avoiding false avoidance of non-critical frequency bands. Furthermore, by statistically analyzing the frequency of resonant excitation states in frequency band x, i.e., the upper limit of the excitation source sensing factor Afrimax(i) ≥ the second resonant excitation threshold F2, and combining this with the sliding time window T to calculate the disturbance density Df(x) of frequency band x, trend analysis of frequency-sensitive areas is achieved, quantifying the intensity of resonant risk.

[0126] Based on the aforementioned dual indicators, a frequency avoidance window function, Wavoid, is constructed to form a decision-making mechanism for determining whether frequency band x should be avoided. Specifically, when the disturbance density Df is higher than 95% of the historical density distribution threshold and the status label Lf indicates that the band has a history of resonance excitation, Wavoid(x) is set to 1, marking it as a zone that must be avoided; otherwise, it is set to 0, representing a freely operable frequency band. This implementation significantly improves the ability to identify frequency bands with potential resonance risks. By abstracting resonance excitation behavior from a single-point event to frequency band behavior, a data-driven sensitive frequency band identification mechanism is constructed, achieving the following objectives and beneficial effects: First, it achieves a leap from transient identification to trend identification; second, it avoids unnecessary frequency adjustment operations, improving system stability; and third, it forms a flexible and scalable strategy foundation, providing structured input for subsequent frequency adjustments, thereby significantly improving the intelligence and precision of system resonance prevention and control.

[0127] Example 6

[0128] Please see Figure 1 Specifically: S5 includes S51;

[0129] S51. Calculate the adjusted target frequency Fadj based on the current avoidance frequency window function Wavoid and the excitation source sensing factor Afri of each water pump, and then transmit the adjusted target frequency Fadj to the frequency converter to adjust the current water pump operating frequency.

[0130] The target frequency Fadj is calculated and output using the following algorithm formula;

[0131] ;

[0132] In the formula, Fadj i (t) represents the target frequency of the i-th water pump at time t, and d represents the integral function. This represents the adjustment response factor, used to control the adjustment speed. It is configured by the user and has a dimensionless value, Wavoid(f). i (t) represents the output frequency f of the i-th water pump at time t. i The avoidance frequency window function of (t), Wavoid(f) i When (t) = 1, it indicates that the output frequency f of the i-th water pump at time t is... i (t) needs to be avoided, Wavoid(f) i When (t) = 0, it indicates that the output frequency f of the i-th water pump at time t is... i (t) No need to avoid;

[0133] The derivation logic of the formula: This value represents the directional adjustment mechanism. It indicates whether Afri will rise or fall, and whether it will rise faster or slower, if the frequency is slightly increased or decreased. If it is positive, increasing the frequency will aggravate the resonance, and the frequency should be adjusted in the opposite direction. If it is negative, increasing the frequency can alleviate the resonance, and the frequency can be adjusted upward. This value provides an adjustment direction similar to the gradient descent method, making the control system more like a sensor.

[0134] This indicates the avoidance adjustment shielding mechanism. This item is used to adjust the shielding of dangerous frequency bands. If the current frequency is already in the avoidance band, Wavoid(f) i When (t) = 1, this term 1 minus 1 becomes 0, and the adjustment range is 0. If the frequency is not marked as the avoidance segment, this term is 1, and then normal adjustment is performed. This forms a mechanism that adjusts when not in the danger zone and jumps when in the danger zone.

[0135] S5 also includes S52;

[0136] S52. After adjusting the pump operating frequency through the target frequency Fadj, the rhythm control and anti-oscillation protection mechanism is executed. The rhythm control and anti-oscillation protection mechanism includes frequency locking, minimum adjustment interval, avoidance trigger conditions and convergence judgment.

[0137] Frequency locking locks the adjusted target frequency Fadj within a range of 1% up and down to prevent excessive frequency jumps from causing secondary disturbances. Even if adjustment is needed, it gradually avoids the resonance zone within a small range.

[0138] The minimum adjustment interval is set to at least 10 seconds between each adjustment operation to prevent frequent frequency jumps from causing system jitter, dead loops, or load fluctuations.

[0139] If the avoidance trigger condition is that the frequency jump is still at level III after three consecutive jumps, the forced avoidance mechanism will be activated, directly limiting the adjustment range to within 5% above and below, and quickly leaving the risk area.

[0140] The convergence judgment is made if the frequency is below Level II for three consecutive cycles. If the frequency is then determined to be effective, the current frequency is added to the safe operating frequency band list, the adjustment is turned off, and the water pump is restored to normal operation.

[0141] In this embodiment, the method constructs a dynamic adjustment and protection control mechanism for the target frequency Fadj, achieving intelligent response to pump frequency adjustment and closed-loop control of oscillation risks. First, based on the excitation source sensing factor Afri of each pump and the avoidance frequency window function Wavoroid to which the current frequency belongs, a directional adjustment mechanism and an avoidance shielding mechanism are introduced to jointly calculate and output the adjusted target frequency Fadj. When the excitation source sensing factor Afri shows an upward trend and the current frequency falls within the frequency range marked as 1 in the avoidance frequency window function Wavoroid, continuous adjustment operations are prohibited in this frequency range. Instead, a frequency jump strategy is executed to quickly avoid high-risk resonance frequency bands. When the current frequency does not belong to the avoidance range, a sensitive fine-tuning is performed based on the response trend of the excitation source sensing factor Afri to the direction of frequency change, making the adjustment direction more targeted and stable. Furthermore, through rhythmic control and protection constraints on the target frequency Fadj, the frequency adjustment operation is prevented from triggering new disturbance sources. Among them, the frequency locking limit is within ±1% of the frequency fluctuation range to avoid the impact of instantaneous large frequency hopping on the system; the minimum adjustment interval strategy ensures that the adjustment period is not less than 10 seconds, effectively eliminating the system oscillation and frequency jitter that may be caused by rapid frequency hopping; the avoidance trigger condition mechanism immediately expands the adjustment range to ±5% when the system still cannot get rid of the Level III resonance level after 3 consecutive frequency jumps, accelerating the move away from the danger zone; and the convergence judgment mechanism automatically closes the adjustment loop and marks the current frequency as a safe frequency band when the system is in a low-risk state for 3 consecutive cycles, thus realizing closed-loop optimization and dynamic self-learning of the adjustment logic.

[0142] Example 7

[0143] Please see Figure 1 and Figure 2 A secondary water supply operation data processing system based on multi-dimensional analysis includes an operation data acquisition module, a frequency excitation source sensing module, a resonance excitation level classification module, an avoidance frequency construction module, and a dynamic frequency adjustment module.

[0144] The operation data acquisition module embeds an API application interface in the central controller of the secondary water supply pump to collect operation data in real time and transmit the operation data to the operation data processing server through the API application interface.

[0145] The frequency excitation source sensing module preprocesses the running data in the running data processing server to obtain a standardized dataset, and calculates and outputs the excitation source sensing factor Afri based on the standardized dataset.

[0146] The resonance excitation level classification module extracts the upper limit value of the excitation source sensing factor Afri based on the excitation source sensing factor Afri. max And set the threshold for the resonance excitation range to conduct a preliminary comparative evaluation;

[0147] The avoidance frequency construction module triggers the avoidance frequency window construction mechanism based on the preliminary comparison and evaluation results, and constructs the avoidance frequency window function Wavoid.

[0148] The dynamic frequency adjustment module performs aggregated calculations based on the excitation source sensing factor Afri and the avoidance frequency window function Wavoroid for each water pump, outputs the adjusted target frequency Fadj, and executes rhythm control and anti-oscillation protection mechanisms.

[0149] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention.

Claims

1. A method for processing secondary water supply operation data based on multi-dimensional analysis, characterized in that: Includes the following steps: S1. By embedding an API application interface in the central controller of the secondary water supply pump, the operating data is collected in real time and transmitted to the operating data processing server through the API application interface. The operational data includes the output frequency f of the i-th water pump at time t. i (t) The current disturbance acceleration AI of the i-th water pump at time t i (t) and the pressure echo delay time Tpr(t) at time t; S2. In the running data processing server, preprocess the running data to obtain a standardized dataset, and perform calculations based on the standardized dataset to output the excitation source perception factor Afri. S2 further includes S22; S22. Start the water pump for secondary water supply. As the water pump's current secondary water supply status changes, continuously update the standardized dataset. At the same time, build a frequency excitation source sensing algorithm model for the water pump in the running data processing server. Input the real-time updated standardized dataset into the frequency excitation source sensing algorithm model for calculation and output the excitation source sensing factor Afri for each water pump to measure the degree of risk of resonance induction by the operating frequency of the water pump during the secondary water supply process. The excitation source sensing factor Afri is calculated and output using the following frequency excitation source sensing algorithm model; ； In the formula, Afri i (t) represents the excitation source sensing factor of the i-th water pump at time t, sin represents the cosine function, π represents pi, with a value of 3.14, and fres represents the center value of the frequency cluster fitting of the water pump resonance event, with a dimensionless value. S3. Extract the upper limit value of the excitation source sensing factor Afri based on the excitation source sensing factor Afri. max And set the threshold for the resonance excitation range to conduct a preliminary comparative evaluation; S4. Based on the preliminary comparison and evaluation results, trigger the avoidance frequency window construction mechanism and construct the avoidance frequency window function Wavoid. S5. Based on the excitation source sensing factor Afri and the avoidance frequency window function Wavoroid of each water pump, perform summary calculations, output the adjusted target frequency Fadj, and execute rhythm control and anti-oscillation protection mechanisms.

2. The method for processing secondary water supply operation data based on multi-dimensional analysis according to claim 1, characterized in that: S1 includes S11 and S12; S11. By embedding an API application interface in the pump center controller of the secondary water supply and deploying an API client in the pump center controller through the API application interface, the pump frequency conversion control terminal, pump motor terminal and resonance response recording module of the pump center controller are automatically connected. After the connection is completed, the API client will automatically run to send a data acquisition command to the water pump central controller to collect the water pump's operating data in the secondary water supply process in real time. The API application interface establishes connections with the PLC control system of the water pump central controller, the ABB water pump smart meter, and the water pump pressure valve. S12. Configure the network access function of the API application interface, use the HTTPS transmission protocol to establish a network connection between the running data processing server and the API application interface, and transmit the running data to the running data processing server over the network.

3. The method for processing secondary water supply operation data based on multi-dimensional analysis according to claim 2, characterized in that: S2 includes S21; S21. Preprocess the running data in the running data processing server to obtain a standardized dataset; The preprocessing includes time series standardization, data cleaning, and normalization. The time series standardization uses a regular sampling rearrangement method to uniformly timestamp the sampling time points of all parameters in the running data; The data cleaning process involves filling in missing values ​​using a sliding window mean when the continuous missing time in the running data is ≤3 seconds, and directly removing the corresponding values ​​if the missing time exceeds 10 seconds.

4. The method for processing secondary water supply operation data based on multi-dimensional analysis according to claim 3, characterized in that: S3 includes S31; S31. Based on the excitation source sensing factor Afri of each water pump, extract the maximum value within a 60-second sliding window to obtain the upper limit value Afri of the excitation source sensing factor. max The upper limit of the excitation source sensing factor Afri max The specific extraction formula is as follows: In the formula, Afri max (i) represents the upper limit of the sensing factor of the excitation source of the i-th water pump, max means taking the upper limit value, and T represents the sliding window time length.

5. The method for processing secondary water supply operation data based on multi-dimensional analysis according to claim 4, characterized in that: S3 includes S32; S32. Extract historically confirmed resonance time samples, analyze the distribution of the excitation source sensing factor Afri in the resonance time samples, and extract the resonance excitation interval threshold, which includes a first resonance excitation threshold F1 and a second resonance excitation threshold F2. Then, use the real-time acquired upper limit value Afri of the excitation source sensing factor of the i-th water pump. max (i) A preliminary comparison and evaluation is conducted with the resonance excitation interval threshold, and the resonance excitation level is classified based on the preliminary comparison and evaluation results to determine the resonance sensitivity state of the current water pump during secondary water supply. Then, based on the resonance excitation level classification results, an avoidance frequency window construction mechanism is triggered. The specific evaluation content is as follows: When the upper limit of the sensing factor of the excitation source of the i-th water pump is Afri max (i) When the frequency is less than the first resonance excitation threshold F1, it indicates a safe frequency band. At this time, it is classified as Level I and no intervention is required for normal operation. When the first resonance excitation threshold F1 ≤ the upper limit value of the sensing factor of the i-th water pump excitation source Afri max (i) When < the second resonance excitation threshold F2, it indicates the secondary excitation frequency band, which is classified as Level II. At this time, the sampling frequency of the API application interface is increased by 50%. When the upper limit of the sensing factor of the excitation source of the i-th water pump is Afri max (i) When the second resonance excitation threshold F2 is greater than or equal to the second resonance excitation threshold, it indicates the resonance excitation frequency band. At this time, it is classified as Level III, triggering the avoidance frequency window construction mechanism.

6. The method for processing secondary water supply operation data based on multi-dimensional analysis according to claim 5, characterized in that: S4 includes S41 and S42; S41. After preliminary comparison and evaluation of the trigger avoidance frequency window construction mechanism, use the output frequency f of the i-th water pump at time t. i (t) Perform discrete binning processing to divide the entire frequency range into several non-overlapping intervals to obtain frequency segment x; Based on the preliminary comparative evaluation of the resonance excitation level classification results, for each frequency band x, the upper limit value of the sensing factor Afri of the excitation source of the i-th water pump is used. max (i) The resonance excitation level classification results are labeled with Level I, Level II and Level III, and the frequency band status label Lf(x) of frequency band x is obtained. Simultaneously, the upper limit value of the sensing factor for the excitation source of the i-th water pump, Afri, was selected. max (i) The time and corresponding frequency of the second resonance excitation threshold F2 are mapped into the frequency segment x. The number of frequencies belonging to the resonance excitation frequency segment on the frequency segment x is counted. The perturbation density Df(x) of the frequency segment x is obtained by dividing by the sliding window time length T. S42. Based on the frequency band status label Lf(x) and the perturbation density Df(x) of frequency band x, construct the avoidance frequency window function Wavoid, and determine the resonance to avoid frequency band x. The specific construction method of the avoidance frequency window function Wavoroid is as follows: ； In the formula, Wavoid(x) represents the avoidance frequency window function for frequency band x, and otherwise means the opposite. The avoidance frequency window function Wavoid=1 means that avoidance is required, and the avoidance frequency window function Wavoid=0 means that no avoidance is required and free operation is allowed. The perturbation density Df(x) of frequency band x > 95% means that the perturbation density Df(x) of frequency band x exceeds 95% of the excitation density, and is judged as a density anomaly.

7. The method for processing secondary water supply operation data based on multi-dimensional analysis according to claim 6, characterized in that: S5 includes S51; S51. Calculate the adjusted target frequency Fadj based on the current avoidance frequency window function Wavoid and the excitation source sensing factor Afri of each water pump, and then transmit the adjusted target frequency Fadj to the frequency converter to adjust the current water pump operating frequency. The target frequency Fadj is calculated and output using the following algorithm formula; ； In the formula, Fadj i (t) represents the target frequency of the i-th water pump at time t, and d represents the integral function. This represents the adjustment response factor, used to control the adjustment speed. It is configured by the user and has a dimensionless value, Wavoid(f). i (t) represents the output frequency f of the i-th water pump at time t. i The avoidance frequency window function of (t).

8. The method for processing secondary water supply operation data based on multi-dimensional analysis according to claim 7, characterized in that: S5 also includes S52; S52. After adjusting the pump operating frequency through the target frequency Fadj, a rhythm control and anti-oscillation protection mechanism is executed. The rhythm control and anti-oscillation protection mechanism includes frequency locking, minimum adjustment interval, avoidance trigger conditions, and convergence judgment. The frequency locking is achieved by locking the adjusted target frequency Fadj, with the adjustment range limited to ±1%. The minimum adjustment interval is defined by limiting the interval between each adjustment operation to at least 10 seconds; The avoidance trigger condition is that if the frequency jump is still at level III after three consecutive frequency jumps, the forced avoidance mechanism will be activated, directly limiting the adjustment range to within 5% above and below, and quickly leaving the risk area. The convergence judgment is made by determining that the adjustment strategy is effective if the frequency is below Level II for three consecutive cycles, adding the current frequency to the safe operation frequency band list, turning off the adjustment, and restoring the water pump to normal operation.

9. A secondary water supply operation data processing system based on multi-dimensional analysis, used to implement the secondary water supply operation data processing method based on multi-dimensional analysis as described in any one of claims 1-8, characterized in that: It includes a data acquisition module, a frequency excitation source sensing module, a resonance excitation level classification module, an avoidance frequency construction module, and a dynamic frequency adjustment module; The operation data acquisition module acquires operation data in real time by embedding an API application interface in the central controller of the secondary water supply pump, and transmits the operation data to the operation data processing server through the API application interface. The frequency excitation source sensing module preprocesses the running data in the running data processing server to obtain a standardized dataset, and calculates based on the standardized dataset to output the excitation source sensing factor Afri. The resonance excitation level classification module extracts the upper limit value of the excitation source sensing factor Afri based on the excitation source sensing factor Afri. max And set the threshold for the resonance excitation range to conduct a preliminary comparative evaluation; The avoidance frequency construction module triggers the avoidance frequency window construction mechanism based on the preliminary comparison and evaluation results, and constructs the avoidance frequency window function Wavoid. The dynamic frequency adjustment module performs a summary calculation based on the excitation source sensing factor Afri and the avoidance frequency window function Wavoroid of each water pump, outputs the adjusted target frequency Fadj, and executes rhythm control and anti-oscillation protection mechanisms.

Images