An intelligent emergency lighting power supply management system based on water pressure prediction and a method thereof

Abstract

The application provides a water pressure prediction-based intelligent emergency lighting power supply management system and method, which can realize prospective management of community riser overpressure water energy and reliable power supply of emergency lighting; the system comprises a water pressure prediction unit for generating a water pressure prediction curve; a hydroelectric power generation unit arranged in an overpressure section of a community water supply riser for generating power by using the riser surplus water pressure; an energy storage management unit for storing electric energy and monitoring the state of charge in real time; an emergency lighting load unit for providing power for community emergency lighting in an emergency state; and an intelligent dispatching controller in communication connection with the water pressure prediction unit, the hydroelectric power generation unit, the energy storage management unit and the emergency lighting load unit, for dynamically adjusting the working mode of the hydroelectric power generation unit and the power generation power distribution strategy according to the water pressure prediction curve and the state of charge of the energy storage management unit, and preferentially guaranteeing the power supply of the emergency lighting load unit when the commercial power is interrupted.

Description

Technical Field

[0001] This invention relates to the field of building water supply energy conservation and emergency power supply technology, specifically to an intelligent emergency lighting power supply management system and method based on water pressure prediction. Background Technology

[0002] High-rise building water supply systems typically employ a vertical zoning system to overcome the problem of excessive static pressure in lower-level pipes. According to relevant regulations on water conservation in civil buildings, the static pressure at the lowest sanitary fixture distribution point in each zone should not exceed 0.45 MPa. Pressure-reducing facilities should be installed in the lower levels of each zone to ensure that the water supply pressure at each water point does not exceed 0.2 MPa. Taking an 18-story high-rise residential building as an example, more than half of the floors require pressure-reducing water supply, resulting in the wasted water energy provided by the variable frequency pump being squandered by the mechanical pressure-reducing valve. To recover this wasted energy, existing technologies have attempted to replace traditional pressure-reducing valves with hydroelectric generators in the water supply system, for example, Chinese patent CN2118. Patent 52038U discloses a high-rise residential water supply overpressure recovery and power generation system, including a generator-type pressure reducing valve, an energy storage device, and a lighting distribution box. The generator-type pressure reducing valve is installed on the water supply branch pipes of each lower floor, and uses water kinetic energy to convert into electrical energy for storage and use in public lighting in corridors. Chinese patent CN106640486A discloses a pressure reducing and power generation system, which uses a pipeline-type power generation device to replace the traditional pressure reducing valve in the water supply system for lighting and other purposes. Chinese patent application CN202211241983 discloses a high-rise building water supply system surplus head recovery system, which adopts a dual-path design and realizes the on / off control of the solenoid valve through a pressure sensor and a control cabinet.

[0003] However, the aforementioned existing technologies have the following shortcomings: First, there is a lack of predictability. Existing systems are all passive power generation, which can only generate electricity under the current water pressure conditions and cannot predict future water pressure changes. The water pressure of the water supply risers in the community has obvious time-varying characteristics (such as large water pressure fluctuations during peak water usage in the morning and evening, and relatively stable water pressure during off-peak water usage at night). Passive power generation cannot make scheduling arrangements in advance according to water pressure changes, resulting in low power generation efficiency or insufficient energy storage.

[0004] Second, the reliability of emergency power supply is insufficient. Although existing technologies store generated electricity in batteries for lighting, they do not consider power priority management in emergency scenarios such as mains power outages. When energy storage is limited, there is a lack of a priority guarantee mechanism for emergency lighting loads.

[0005] Third, there is a lack of intelligent scheduling capabilities. Current technologies lack coordinated optimization among power generation, energy storage, and power consumption, and cannot dynamically adjust power generation and distribution strategies based on water pressure prediction results and energy storage status.

[0006] Fourth, the efficiency of hydropower utilization is not high. Existing hydropower generation devices typically operate under fixed conditions and cannot adjust their operating point according to real-time water pressure and flow rate to maintain optimal efficiency. Summary of the Invention

[0007] To address the aforementioned problems, this invention provides an intelligent emergency lighting power supply management system and method based on water pressure prediction, which can realize the forward-looking management of overpressure water energy in the riser pipes of the community and the reliable power supply for emergency lighting.

[0008] This invention adopts the following technical solution: an intelligent emergency lighting power supply management system based on water pressure prediction, comprising a water pressure prediction unit, a hydroelectric power generation unit, an energy storage management unit, an emergency lighting load unit, and an intelligent dispatch controller; wherein, The water pressure prediction unit is used to collect historical water pressure data, water usage period characteristics, and building water usage patterns of the water supply risers in the community. It uses a time-series prediction model to predict the riser water pressure within a preset time window and generates a water pressure prediction curve. The hydroelectric power generation unit is installed in the overpressure section of the water supply riser in the community to generate electricity using the excess water pressure in the riser. An energy storage management unit, electrically connected to the hydropower generation unit, is used to store electrical energy and monitor the state of charge in real time; An emergency lighting load unit, electrically connected to the energy storage management unit, is used to provide power for the community's emergency lighting in an emergency situation. The intelligent dispatch controller is communicatively connected to the water pressure prediction unit, the hydropower generation unit, the energy storage management unit, and the emergency lighting load unit, respectively. It is used to dynamically adjust the working mode and power distribution strategy of the hydropower generation unit according to the water pressure prediction curve and the state of charge of the energy storage management unit, and to prioritize the power supply of the emergency lighting load unit when the mains power is interrupted.

[0009] Furthermore, the intelligent scheduling controller includes: The prediction and evaluation module is used to evaluate the range of generateable energy within a future time window based on the water pressure prediction curve. An energy storage status monitoring module is used to acquire the state of charge and health status of the energy storage management unit in real time. The priority decision module has a built-in multi-level load priority scheduling strategy, and sets the emergency lighting load as the highest priority; The scheduling and execution module is used to allocate surplus power to non-critical loads or integrate it into the community microgrid, based on the prediction and assessment results and the energy storage charge status, while meeting the backup power needs for emergency lighting. The emergency switching module is used to obtain the available time of emergency lighting based on the water pressure prediction curve and the energy storage charge status when the mains power is interrupted, and to issue an early warning signal when the energy storage is insufficient. Furthermore, the priority decision module incorporates a three-level load priority system, divided into first priority, second priority, and third priority, with scheduling strategies for each level as follows: The first priority is emergency lighting load, and the backup power capacity of the energy storage management unit is continuously maintained at no less than the preset capacity threshold. The second priority is non-emergency lighting and security monitoring load in public areas; The third priority is the regional loads that can be interrupted with flexible control and the microgrid feeder loads. Furthermore, the preset capacity threshold for emergency lighting load is set to 80% of the rated energy storage capacity of the energy storage management unit; Furthermore, the hydropower generation unit is a modular bypass pipeline architecture, which includes an independently set main water supply path and a bypass power generation path. The main water supply path and the bypass power generation path are connected in parallel to the overpressure section between the community water supply riser and the downstream riser. A pressure reducing valve is fixedly installed on the main water supply path, and a pipeline turbine generator and a flow regulating valve are sequentially installed on the bypass power generation path. Furthermore, the hydroelectric power generation unit also includes: The rectifier and voltage regulator module is electrically connected to the pipeline turbine generator and is used to rectify the alternating current output by the pipeline turbine generator, convert the AC power into DC power, and complete the voltage regulation. The maximum power point tracking module is electrically coupled to the pipeline turbine generator and is used to adjust the load of the pipeline turbine generator according to the real-time water pressure and flow rate, so that the hydropower generation unit operates at the optimal efficiency point under different water pressure conditions. Furthermore, the water pressure prediction unit includes: The data acquisition module is used to collect riser inlet pressure, outlet pressure, instantaneous flow rate, and timestamp data in real time. A data preprocessing module, connected to the data acquisition module, is used to clean, filter, and remove outliers from the acquired data. The feature extraction module, connected to the data preprocessing module, is used to extract water usage time period features, periodic patterns, and pressure fluctuation features from the preprocessed data. The prediction engine module, which has a built-in time-series prediction model, is connected to the feature extraction module and is used to output the water pressure prediction curve within a preset time window based on historical data and feature parameters. Furthermore, the system also includes a communication unit for uploading the operating status, power generation data, energy storage data and prediction results of the intelligent dispatch controller to a cloud management platform or a community property management system. The human-machine interaction unit is used to display operating status, prediction curves, energy storage capacity, and emergency lighting backup power duration information. A method for intelligent emergency lighting power supply management based on water pressure prediction includes the following steps: S1. Collect historical water pressure data, water usage time characteristics, and building water usage patterns of the community's water supply risers; S2. Use a time-series prediction model to predict the riser water pressure within a preset time window and generate a water pressure prediction curve; S3. Real-time monitoring of the state of charge of the energy storage management unit; S4. Based on the water pressure prediction curve and the state of charge, dynamically adjust the working mode and power distribution strategy of the hydropower generation unit. S5. When the mains power is interrupted, based on the water pressure prediction curve and the current energy storage status, obtain the available time of emergency lighting, and allocate power according to the multi-level load priority scheduling strategy to give priority to the emergency lighting load. S6. If a period of low water pressure is predicted in the future based on the water pressure prediction curve and the energy storage is insufficient, increase the energy storage charging power or issue a maintenance warning signal in advance before the low water pressure arrives.

[0010] Further, in step S4, the dynamic adjustment includes: When it is predicted that the water pressure will be higher than the first preset threshold within the next T time period, the hydropower generation unit will be controlled to generate electricity at the rated power, and the surplus electrical energy will be stored in the energy storage management unit and allocated to non-critical loads. When it is predicted that the water pressure will be between the first and second preset thresholds within the next T time period, the hydropower generation unit will be controlled to operate at the optimal efficiency point, and the generated electricity will be prioritized for energy storage charging. When it is predicted that the water pressure will be lower than the second preset threshold within a future time period T, the hydropower generation unit will reduce its power generation or suspend power generation, and the energy storage management unit will maintain the power supply for emergency lighting loads.

[0011] The beneficial effects of this invention are that, through the water pressure prediction unit, hydropower generation unit, energy storage management unit, emergency lighting load unit, and intelligent dispatch controller, the future water pressure change trend can be accurately predicted, thereby completing the forward-looking optimization of the power generation strategy, completely avoiding the lag problem of the traditional passive power generation mode, and combining water pressure prediction data to complete the power reserve in advance, effectively ensuring the power supply stability and reliability of emergency lighting in the event of a mains power outage. At the same time, it can also realize functions such as remote monitoring of equipment operation status and intelligent analysis of operation data, greatly reducing the difficulty of system operation and maintenance, facilitating subsequent system function iteration and continuous performance optimization, and possessing strong scenario adaptability and scalability. Attached Figure Description

[0012] Figure 1This is a block diagram of the overall architecture of the intelligent emergency lighting power supply management system based on water pressure prediction of the present invention; Figure 2 This is a structural block diagram of the water pressure prediction unit of the present invention; Figure 3 This is a structural block diagram of the intelligent scheduling controller of the present invention; Figure 4 This is a schematic diagram of the modular bypass pipeline structure of the hydropower generation unit of the present invention; Figure 5 This is a flowchart illustrating the intelligent emergency lighting power supply management method based on water pressure prediction according to the present invention. Figure 6 This is a schematic diagram illustrating how the intelligent dispatch controller dynamically adjusts the power generation strategy based on the water pressure prediction curve in this invention. Detailed Implementation

[0013] like Figures 1-6 As shown, the present invention discloses an intelligent emergency lighting power supply management system based on water pressure prediction, comprising a water pressure prediction unit, a hydroelectric power generation unit, an energy storage management unit, an emergency lighting load unit, and an intelligent dispatch controller; wherein, The water pressure prediction unit is used to collect historical water pressure data, water usage period characteristics, and building water usage patterns of the water supply risers in the community. It uses a time-series prediction model to predict the riser water pressure within a preset time window and generates a water pressure prediction curve. Preferably, the time-series prediction model is a deep learning prediction model based on Long Short-Term Memory (LSTM) or Gated Recurrent Unit (GRU). This model takes historical water pressure data, flow data, and time segment labels for N consecutive time steps as input and outputs the water pressure change trend within the next time window. Before deployment, the time-series prediction model is trained and validated using historical water usage data of the community. For example, if an existing LSTM neural network prediction model based on time-series big data is used, the historical 7×24h water pressure and flow time-series data of the community are pre-trained and recorded. Combined with the water usage patterns of different time periods, seasonal water usage characteristics, and the fluctuation characteristics of morning and evening water usage peaks, the model dynamically predicts the water pressure and flow change trends of the pipeline network in the next 1h, 3h, and 6h, and outputs the water pressure fluctuation range, peak water pressure, low water pressure, and corresponding flow parameters, providing data support for the system's advance scheduling. The hydroelectric power generation unit is installed in the overpressure section of the water supply riser in the community to generate electricity using the excess water pressure in the riser. The energy storage management unit is electrically connected to the hydropower generation unit and is used to store electrical energy and monitor the state of charge (SOC) and state of health (SOH) in real time. The emergency lighting load unit is electrically connected to the energy storage management unit and is used to provide power for the community's emergency lighting in an emergency situation. The intelligent dispatch controller is connected to the water pressure prediction unit, the hydropower generation unit, the energy storage management unit, and the emergency lighting load unit. It is used to dynamically adjust the working mode and power distribution strategy of the hydropower generation unit according to the water pressure prediction curve and the state of charge of the energy storage management unit, and to prioritize the power supply of the emergency lighting load unit when the mains power is interrupted.

[0014] The intelligent dispatch controller includes: The prediction and assessment module is used to assess the range of power-generating energy within a future time window based on the water pressure prediction curve. The energy storage status monitoring module is used to acquire the state of charge and health status of the energy storage management unit in real time. The priority decision module has a built-in multi-level load priority scheduling strategy and sets the emergency lighting load as the highest priority; The scheduling and execution module is used to allocate surplus power to non-critical loads or integrate it into the community microgrid, based on the prediction and assessment results and the energy storage charge status, while meeting the backup power needs for emergency lighting. The emergency switching module is used to obtain the available time of emergency lighting based on the water pressure prediction curve and the energy storage charge status when the mains power is interrupted, and to issue an early warning signal when the energy storage is insufficient. The priority decision module has a built-in three-level load priority system, divided into first priority, second priority, and third priority, and the scheduling strategies for the three levels of load priority are as follows: The first priority is emergency lighting load, and the backup power capacity of the energy storage management unit must be maintained at no less than the preset capacity threshold. The preset capacity threshold for emergency lighting load is set at 80% of the rated energy storage capacity of the energy storage management unit. The second priority is non-emergency lighting and security monitoring load in public areas; The third priority is the regional loads that can be interrupted with flexible control and the microgrid feeder loads. Specifically, this patented intelligent emergency lighting power supply management system always prioritizes the emergency lighting loads in the first priority category. Only when the power exceeds 80% of the rated capacity of the energy storage can the excess power (i.e., surplus power) be used to meet the needs of the second and third priorities. Once the power drops below 80% of the rated capacity of the energy storage, the second and third priority loads are immediately "cut off" and all power is given to the emergency lighting loads in the first priority category.

[0015] The hydropower generation unit is a modular bypass pipeline architecture. The modular bypass pipeline architecture includes an independently set main water supply channel and a bypass power generation channel. The main water supply channel and the bypass power generation channel are connected in parallel to the overpressure section between the community water supply riser and the downstream riser. A pressure reducing valve is fixedly installed on the main water supply channel, and a pipeline turbine generator and a flow regulating valve are installed in sequence on the bypass power generation channel. Among them, the pipeline-type water turbine generator is installed in the overpressure section of the water supply riser in the community. It can collect the pressure energy and kinetic energy of the water flow in the pipeline and convert it into electrical energy, so as to realize the recovery and utilization of the residual pressure water energy in the pipeline. Under normal water supply conditions, the water flows through the bypass power generation path in a controlled manner, and generates hydropower through the pipeline turbine generator. At the same time, the flow rate of the water in the bypass power generation path is precisely controlled by the flow regulating valve. When the pipeline turbine generator needs to be inspected and maintained or when the pipeline experiences abnormal water pressure, the pipeline water flow can be automatically switched to the main water supply path. The pressure-stabilizing water supply is completed through the pressure-reducing valve of the main water supply path, without interrupting the overall water supply of the community. This effectively ensures the continuity and stability of the water supply system and ensures that water supply safety is not affected. The hydroelectric power generation unit also includes: The rectifier and voltage regulator module is electrically connected to the pipeline turbine generator. It is used to rectify the alternating current output by the pipeline turbine generator, convert the AC power into DC power, and perform voltage regulation to ensure the stability of the output power. The maximum power point tracking module is electrically coupled to the pipeline turbine generator. It can collect dynamic parameters of pipeline water pressure and flow rate in real time, and adaptively adjust the load conditions of the pipeline turbine generator based on real-time operating parameters. This ensures that the hydropower generation unit operates at the optimal efficiency point under different water pressure and flow conditions, that is, it can maintain the optimal power generation efficiency and adapt to the working characteristics of dynamic fluctuations in water pressure in the community water supply pipeline.

[0016] The water pressure prediction unit includes: The data acquisition module is used to collect riser inlet pressure, outlet pressure, instantaneous flow rate, and timestamp data in real time. The data preprocessing module, connected to the data acquisition module, is used to clean, filter, and remove outliers from the acquired data. The feature extraction module, connected to the data preprocessing module, is used to extract water usage time characteristics, periodic patterns, and pressure fluctuation characteristics from the preprocessed data. The prediction engine module, which has a built-in time series prediction model, is connected to the feature extraction module and is used to output the water pressure prediction curve within a preset time window based on historical data and feature parameters.

[0017] The water pressure prediction curve is generated by the prediction engine module in the water pressure prediction unit, which includes the following six steps: Step 1: Data Collection The data acquisition module collects real-time operational data of the community's water supply risers, forming a raw time-series dataset; The collected data includes:

[0018] The preferred data acquisition frequency is once every 1 minute to capture fine-grained variations in water use behavior. The acquired data are arranged in chronological order to form a multivariate time series sample set. X(t) = {Pin(t), Pout(t), Q(t)}; Where X(t) represents the multivariate observation vector at time t; Step 2: Data Preprocessing The data preprocessing module cleans, filters, and removes outliers from the raw collected data to improve data quality.

[0019] (1) Handling missing values For data loss caused by momentary sensor malfunction, linear interpolation is used to fill in the gaps. x missing = x t-1 +(x t+1 -x t-1 ) / 2; Where, x missing Represented as the estimated value of the data at the time of the missing data; x t-1 This represents the data value at the time preceding the missing time. x t+1 This represents the data value at the time immediately following the missing time. (2) Outlier removal The 3σ criterion (Raida criterion) is used to identify and remove outliers. For a data sequence {x1,x2,...,xn}, the mean μ and standard deviation σ are calculated.

[0020]

[0021] When a data point satisfies |xi-μ|>3σ, it is identified as an outlier and removed. Linear interpolation is then used to fill in the missing data. Where xi is the i-th data point in the data sequence; n is the total number of data points in the data sequence; μ is the arithmetic mean of the data sequence; σ is the standard deviation of the data sequence; (3) Data normalization To eliminate the influence of different units on model training, minmax normalization is used to scale each feature data to the [0,1] interval:

[0022] in, These are the normalized data values; The original data value; This is the minimum value of the feature in the training dataset; This represents the maximum value of the feature in the training dataset. After normalization, all pressure and flow data are on the same order of magnitude, avoiding gradient vanishing or gradient exploding problems during model training.

[0023] Step 3: Feature Engineering and Sliding Window Construction The feature extraction module extracts time-series features related to water usage patterns from the preprocessed data and constructs supervised learning samples using the sliding window method.

[0024] (1) Temporal feature extraction Based on the periodic patterns of water usage behavior in the community, the following temporal features were extracted:

[0025] (2) Sliding window construction A sliding window method is used to convert continuous time series into supervised learning samples. With a lookback window length L (i.e., using data from the past L time steps to predict the next T time steps), the k-th training sample is constructed as follows:

[0026]

[0027] Where L is the length of the lookback window (input time steps), which is the number of historical data points used for prediction; T represents the prediction window length (output time steps), which is the number of future data points to be predicted. k is the index of the training sample; Let be the input matrix for the k-th training sample; Let be the output vector (target value) of the kth training sample; The start time of the kth training sample; The riser inlet water pressure at time t (the target variable for prediction). Preferably, the length of the retrospective window L = 60 (i.e., the data from the past hour) and the length of the prediction window T = 30 (i.e., the water pressure prediction curve for the next 30 minutes).

[0028] Step 4: Building a Time Series Prediction Model The prediction engine module incorporates a deep learning time series prediction model based on Long Short-Term Memory (LSTM) networks or Gated Recurrent Units (GRUs). Taking LSTM as an example, its core computational unit contains three gating structures.

[0029] (1) Forget Gate The forget gate determines how much information from the cell state ct-1 at the previous time step is retained in the current time step: ft = σ(Wf·[ht-1,Xt] + bf); ft is the forget gate output value at time t, with a value range of [0,1]. σ is the Sigmoid activation function, σ(z) = 1 / (1+ez); Wf is the weight matrix of the forget gate (learnable parameters); ht-1 is the hidden state vector at time t-1; Xt is the input vector at time t (normalized water pressure and flow rate data). bf is the bias term (learnable parameter) of the forget gate; (2) Input Gate The input gate determines how much of the current input information Xt is stored in the cell state: it = σ(Wi·[ht-1,Xt] + bi)

[0030] Where it is the input gate output value at time t, and its value ranges from [0,1]. bi is the bias term of the input gate (a learnable parameter); Let be the candidate cell state vector at time t; The hyperbolic tangent activation function is used. ; The weight matrix for candidate cell states (learnable parameters); Bias terms for candidate cell states (learnable parameters); (3) Cell State Update Update the cell state by combining the outputs of the forget gate and the input gate: ct = ft⊙ct-1 + it⊙ Where ct is the cell state vector (long-term memory) at time t. ⊙ represents element-wise multiplication (Hadamard product); ct-1 is the cell state vector at time t-1; (4) Output Gate The output gate determines how much information in the cell state ct is output as the hidden state ht: ot = σ(Wo·[ht-1,Xt] + bo) ht=ot⊙tanh(ct) Where ot is the output value of the output gate at time t, and its value ranges from [0,1]. Wo is the weight matrix of the output gate (learnable parameters); bo is the bias term of the output gate (a learnable parameter); ht is the hidden state vector (short-term memory) at time t. (5) Fully connected output layer The output ht of the LSTM layer is mapped to the final predicted value through a fully connected layer:

[0031] in, The predicted water pressure at time t (normalized prediction result); Wfc is the weight matrix of the fully connected layer (learnable parameters); bfc is the bias term (learnable parameter) of the fully connected layer; Step 5: Model Training The model needs to be trained and validated using the historical water pressure data of the community before deployment.

[0032] (1) Training dataset partitioning The preprocessed historical dataset was divided into three subsets in chronological order:

[0033] (2) Loss function Mean Squared Error (MSE) is used as the loss function for model training:

[0034] This represents the loss function value (mean squared error). The total number of training samples; The normalized value of the true water pressure of the i-th sample. Predict the water pressure value (after normalization) for the i-th sample using the model. (3) Optimization algorithm The Adam optimizer (Adaptive Moment Estimation) is used to update the gradient of the model parameters, with an initial learning rate of 10. 3 Set the batch size to 32.

[0035] Step Six: Generation of Water Pressure Prediction Curve The trained model is deployed in the prediction engine module, which receives real-time collected data and outputs prediction results.

[0036] (1) Predicting Input Construction In the real-time prediction phase, the input sequence is constructed using historical data from the current time t0 and the previous L1 times: Xpredict=[X(t0-L+1),X(t0-L+2),...,X(t0)] Xpredict is the input matrix for real-time prediction; t0 is the current time (prediction start time); L is the length of the viewport window; (2) Model reasoning Xpredict is input into the trained LSTM model for forward inference, and the model outputs a sequence of predicted water pressure values ​​within the prediction window:

[0037] (3) Anti-normalization Convert the normalized predicted values ​​from the model output back to the actual water pressure values:

[0038] in, The actual predicted water pressure at time t (unit: MPa); This is the normalized predicted water pressure value at time t (model output); The maximum water pressure (unit: MPa) in the training dataset; The minimum water pressure (unit: MPa) in the training dataset; (4) Predicted curve output The final water pressure prediction curve is as follows: Ppred={Ppred(t0+1),Ppred(t0+2),...,Ppred(t0+T)} Wherein, Ppred is the water pressure prediction curve vector, containing the water pressure prediction values ​​at the next T time points; t0 is the current time (prediction start time); T is the prediction window length (output time steps); The water pressure prediction curve is output to the intelligent dispatch controller for forward-looking power generation strategy optimization and energy storage dispatch decisions.

[0039] The intelligent dispatch controller also uses a prediction and evaluation module to assess the range of generateable energy within a future time window based on the water pressure prediction curve. This includes the following steps: First step: Determine the effective head for power generation The fundamental driving force of hydropower generation is the excess pressure energy in the riser, which is the difference between the inlet pressure and the outlet pressure (or the minimum service pressure required by the user).

[0040] (1) Calculation of predicted pressure difference Based on the water pressure prediction curve, calculate the predicted pressure difference between the front and rear ends of the hydro-generator unit:

[0041] in, For a moment The predicted pressure difference (effective power generation pressure) is expressed in MPa. For a moment The predicted inlet water pressure of the hydro-generator unit, in MPa; For a moment The predicted water pressure at the outlet of the hydro-generator unit is expressed in MPa. To predict the i-th time point on the curve, i=1,2,...,T; When the outlet pressure sensor is not directly installed, the outlet pressure can be taken as the minimum service pressure Pmin_service required by the most unfavorable water use point in the community water supply (usually 0.15~0.20 MPa), then: ΔPpred(ti)=Pin_pred(ti)-Pmin_service Pmin_service is the minimum service pressure requirement for the user side (determined by building codes); (2) Conversion of effective power generation head Convert the pressure difference into an equivalent head height:

[0042] The effective hydroelectric head at time ti is expressed in meters. Let be the density of water, taken as 1000-1000 at room temperature; For gravitational acceleration, take 9.81 or 9.81. Step 2: Predicting and estimating flow rates Hydropower generation capacity depends not only on the water head but also on the flow rate. The prediction and assessment module needs to estimate the flow rate within the prediction window.

[0043] (1) Traffic flow period feature mapping Based on the historical water usage data of this community, an average flow feature database for different time periods (hours / weekdays / holidays) is established. For any time point \(t_i\) within the prediction window, the corresponding predicted flow is matched according to its time period label: Qpred(ti=Q(s(ti),d(ti)) Where Qpred(ti) is the predicted instantaneous flow rate at time ti; Q(s,d) represents the historical average flow under time period label s and date type d in the feature library; s(ti) is the time period label (peak / off-peak / valley) for time ti; d(ti) represents the date type (weekday / weekend / holiday) for time ti; (2) Real-time flow correction When a real-time flow sensor is installed in the system, a first-order exponential smoothing method is used to correct the predicted flow in real time. Qpred(ti)=α·Qreal(t0)+(1-α)·Q(s(ti),d(ti)) Where α is the slip coefficient, with a value of 0 < α < 10 < α < 1, and preferably α = 0.7. Qreal(t0) is the real-time measured flow rate at the current time t0; t0 is the current time (prediction start time); Third step: Calculate instantaneous power generation Calculate the instantaneous power generation of the hydro-turbine generator unit at time t_i based on the head and flow rate.

[0044] (1) Hydropower (Input Power) The total power carried by the water flow (water power) is: Pwater(ti)=ρ·g·Qpred(ti)·Hpred(ti) Where Pwater(ti) is the water energy power (total water flow power) at time ti; ρ is the density of water (1000 kg / m³). 3 ); g is the acceleration due to gravity (9.81 m / s²). 2 ); Qpred(ti) is the predicted instantaneous flow rate at time ti; Hpred(ti) is the effective head for generating electricity at time ti; (2) Power generation capacity (output power) Considering the overall efficiency of the hydro-generator unit, the actual power output is: Pgen(ti)=ηtotal(ti)·Pwater(ti) The overall efficiency is:

[0045] Optimized efficiency point: For accurate calculation, \(\eta_{turbine}\) is not a fixed value, but a bivariate function of head and flow rate. The MPPT module ensures that the turbine operates at its optimal efficiency point under various operating conditions.

[0046] fopt(·) is the optimal efficiency interpolation function for the hydro-generator (calibrated by the manufacturer's efficiency curve). Step 4: Calculate the cumulative renewable energy within the window. The total generateable energy is obtained by integrating the instantaneous power generation at all times within the prediction window (discrete summation).

[0047] (1) Discrete summation formula

[0048]

[0049] (2) Unit conversion to commonly used units In engineering, kilowatt-hours (kWh) are commonly used to represent electrical energy. The conversion formula is as follows:

[0050] Total renewable energy, expressed in kilowatt-hours. Step 5: Calculate the range of generateable energy (range estimation) Because water pressure prediction has a certain degree of uncertainty (prediction error), and flow rate also has random fluctuations, the energy that can be generated is output in the form of a range (rather than a single value), providing a more robust decision-making basis for the intelligent dispatch controller.

[0051] (1) Prediction error statistics Based on the model's historical prediction performance on the validation set, the error distribution of the predicted water pressure is statistically analyzed. Definition:

[0052]

[0053] Simultaneously calculate the half-width of the confidence interval for the predicted water pressure:

[0054]

[0055]

[0056] (2) Calculation of upper and lower bounds of energy range By propagating the water pressure prediction error into the energy calculation, the range of generateable energy is obtained: Lower bound of energy (conservative estimate):

[0057] in:

[0058] Energy upper bound (optimistic estimate):

[0059] in:

[0060]

[0061] (3) Final output range The prediction and assessment module ultimately outputs the range of generateable energy:

[0062] Simultaneously output the expected value (optimal estimate):

[0063] in, This is a vector representing the range of generateable energy. The expected value of the generateable energy (midpoint of the interval); Step 6: Output power generation curve (for dispatching use) In addition to the total energy range, the prediction and evaluation module also outputs the power generation curves for each time step within the entire prediction window to the scheduling execution module to support refined charge and discharge scheduling.

[0064] And the corresponding upper and lower envelopes of schedulable power:

[0065]

[0066]

[0067] in, To predict the power generation curve within the prediction window; This is the lower envelope (conservative curve) of the power that can be generated. The upper envelope (optimistic curve) of the generateable power. The relative error coefficient (dimensionless) for power generation prediction. This represents the average value of the water pressure prediction curve within the window.

[0068] The system also includes a communication unit, which is used to upload the operating status of the intelligent dispatch controller, power generation data, energy storage data and forecast results to the cloud management platform or the community property management system; The human-machine interface unit is used to display operating status, prediction curves, energy storage capacity, and emergency lighting backup power duration information.

[0069] A method for intelligent emergency lighting power supply management based on water pressure prediction includes the following steps: S1. Collect historical water pressure data, water usage time characteristics, and building water usage patterns of the community's water supply risers; S2. Use a time-series prediction model to predict the riser water pressure within a preset time window and generate a water pressure prediction curve; S3. Real-time monitoring of the state of charge of the energy storage management unit; S4. Based on the water pressure prediction curve and state of charge, dynamically adjust the working mode and power distribution strategy of the hydropower generation unit. Further, in step S4, as Figure 6 As shown, dynamic adjustments include: If the water pressure is predicted to be higher than a first preset threshold within a future time period T. At the same time, the hydropower generation unit is controlled to generate electricity at its rated power, and the surplus electrical energy is stored in the energy storage management unit and allocated to non-critical loads; When the water pressure is predicted to be within a certain time frame (T), it should be within the first preset threshold. Second preset threshold During this period, the hydropower generation unit is controlled to operate at its optimal efficiency point, and the generated electricity is prioritized for energy storage charging. If the water pressure is predicted to be lower than the second preset threshold within a future time period T. In case of emergency, the hydropower generation unit reduces or suspends power generation, while the energy storage management unit maintains power supply for emergency lighting loads. as well as Figure 6 In the diagram, the vertical dashed line represents the decision trigger point, which is the moment when the system needs to trigger a decision to reduce power or suspend power generation in advance when it is predicted that a period of low water pressure will soon arrive. This is the initiation time of the advance decision, or the decision time point.

[0070] S5. When the mains power is interrupted, the available time of emergency lighting is obtained based on the water pressure prediction curve and the current energy storage status, and power is allocated according to the multi-level load priority scheduling strategy to give priority to the emergency lighting load. S6. If a period of low water pressure is predicted in the future based on the water pressure prediction curve and the energy storage is insufficient, increase the energy storage charging power or issue a maintenance warning signal in advance before the arrival of low water pressure (that is, the time interval of continuous low pressure).

[0071] Compared with the prior art, the present invention has the following beneficial effects: 1. Proactive scheduling: By introducing a water pressure prediction unit, future water pressure change trends can be predicted in advance, enabling proactive optimization of power generation strategies and avoiding the lag defects of passive power generation; 2. High reliability of emergency power supply: Emergency lighting is set as the highest priority through a multi-level load priority scheduling strategy, and power is stored in advance based on water pressure prediction results to ensure reliable power supply for emergency lighting when the mains power is interrupted; 3. High hydropower utilization efficiency: Through the coordinated optimization of the maximum power point tracking module and the intelligent dispatch controller, the hydropower generation unit operates at the optimal efficiency point under different water pressure conditions, thereby improving the hydropower utilization rate; 4. Water supply safety is guaranteed: The modular bypass pipeline design allows water supply to be switched to the main channel pressure reducing valve when the hydropower unit is under maintenance or malfunctions, without affecting normal water supply; 5. The system has strong scalability: it can be connected to the cloud management platform or the community property management system through the communication unit to realize remote monitoring and data analysis, which facilitates system operation and maintenance and continuous optimization.

[0072] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0073] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. An intelligent emergency lighting power supply management system based on water pressure prediction, characterized in that: It includes a water pressure prediction unit, a hydropower generation unit, an energy storage management unit, an emergency lighting load unit, and an intelligent dispatch controller; among which, The water pressure prediction unit is used to collect historical water pressure data, water usage period characteristics, and building water usage patterns of the water supply risers in the community. It uses a time-series prediction model to predict the riser water pressure within a preset time window and generates a water pressure prediction curve. The hydroelectric power generation unit is installed in the overpressure section of the water supply riser in the community to generate electricity using the excess water pressure in the riser. An energy storage management unit, electrically connected to the hydropower generation unit, is used to store electrical energy and monitor the state of charge in real time; An emergency lighting load unit, electrically connected to the energy storage management unit, is used to provide power for the community's emergency lighting in an emergency situation. The intelligent dispatch controller is communicatively connected to the water pressure prediction unit, the hydropower generation unit, the energy storage management unit, and the emergency lighting load unit, respectively. It is used to dynamically adjust the working mode and power distribution strategy of the hydropower generation unit according to the water pressure prediction curve and the state of charge of the energy storage management unit, and to prioritize the power supply of the emergency lighting load unit when the mains power is interrupted.

2. The intelligent emergency lighting power supply management system based on water pressure prediction according to claim 1, characterized in that: The intelligent scheduling controller includes: The prediction and evaluation module is used to evaluate the range of generateable energy within a future time window based on the water pressure prediction curve. An energy storage status monitoring module is used to acquire the state of charge and health status of the energy storage management unit in real time. The priority decision module has a built-in multi-level load priority scheduling strategy, and sets the emergency lighting load as the highest priority; The scheduling and execution module is used to allocate surplus power to non-critical loads or integrate it into the community microgrid, based on the prediction and assessment results and the energy storage charge status, while meeting the backup power needs for emergency lighting. The emergency switching module is used to determine the available duration of emergency lighting based on the water pressure prediction curve and the energy storage charge status when the mains power is interrupted, and to issue an early warning signal when the energy storage is insufficient.

3. The intelligent emergency lighting power supply management system based on water pressure prediction according to claim 2, characterized in that: The priority decision module has a built-in three-level load priority system, divided into first priority, second priority, and third priority, and the scheduling strategies for the three levels of load priority are as follows: The first priority is emergency lighting load, and the backup power capacity of the energy storage management unit is continuously maintained at no less than the preset capacity threshold. The second priority is non-emergency lighting and security monitoring load in public areas; The third priority is the regional loads that can be interrupted with flexible control and the microgrid feeder loads.

4. The intelligent emergency lighting power supply management system based on water pressure prediction according to claim 3, characterized in that: The preset capacity threshold for emergency lighting load is set to 80% of the rated energy storage capacity of the energy storage management unit.

5. The intelligent emergency lighting power supply management system based on water pressure prediction according to claim 1, characterized in that: The hydropower generation unit is a modular bypass pipeline architecture, which includes an independently set main water supply passage and a bypass power generation passage. The main water supply passage and the bypass power generation passage are connected in parallel to the overpressure section between the community water supply riser and the downstream riser. A pressure reducing valve is fixedly installed on the main water supply passage, and a pipeline turbine generator and a flow regulating valve are sequentially installed on the bypass power generation passage.

6. The intelligent emergency lighting power supply management system based on water pressure prediction according to claim 5, characterized in that: The hydroelectric power generation unit also includes: The rectifier and voltage regulator module is electrically connected to the pipeline turbine generator and is used to rectify the alternating current output by the pipeline turbine generator, convert the AC power into DC power, and perform voltage regulation. The maximum power point tracking module is electrically coupled to the inline hydro-generator and is used to adjust the load of the inline hydro-generator according to the real-time water pressure and flow rate, so that the hydropower generation unit operates at the optimal efficiency point under different water pressure conditions.

7. The intelligent emergency lighting power supply management system based on water pressure prediction according to claim 1, characterized in that: The water pressure prediction unit includes: The data acquisition module is used to collect real-time data on the inlet and outlet pressures, instantaneous flow rate, and timestamps of the water supply risers in the community. A data preprocessing module, connected to the data acquisition module, is used to clean, filter, and remove outliers from the acquired data. The feature extraction module, connected to the data preprocessing module, is used to extract water usage time period features, periodic patterns, and pressure fluctuation features from the preprocessed data. The prediction engine module, which has a built-in time-series prediction model, is connected to the feature extraction module and is used to output a water pressure prediction curve within a preset time window based on historical data and feature parameters.

8. The intelligent emergency lighting power supply management system based on water pressure prediction according to claim 1, characterized in that: The system also includes a communication unit for uploading the operating status, power generation data, energy storage data and prediction results of the intelligent dispatch controller to a cloud management platform or a community property management system. The human-machine interface unit is used to display operating status, prediction curves, energy storage capacity, and emergency lighting backup power duration information.

9. A method for intelligent emergency lighting power supply management based on water pressure prediction, employing the intelligent emergency lighting power supply management system based on water pressure prediction as described in any one of claims 1 to 8, characterized in that: The method includes the following steps: S1. Collect historical water pressure data, water usage time characteristics, and building water usage patterns of the community's water supply risers; S2. Use a time-series prediction model to predict the riser water pressure within a preset time window and generate a water pressure prediction curve; S3. Real-time monitoring of the state of charge of the energy storage management unit; S4. Based on the water pressure prediction curve and the state of charge, dynamically adjust the working mode and power distribution strategy of the hydropower generation unit. S5. When the mains power is interrupted, based on the water pressure prediction curve and the current energy storage status, obtain the available time of emergency lighting, and allocate power according to the multi-level load priority scheduling strategy to give priority to the emergency lighting load. S6. If a period of low water pressure is predicted in the future based on the water pressure prediction curve and the energy storage is insufficient, increase the energy storage charging power or issue a maintenance warning signal in advance before the low water pressure arrives.

10. The intelligent emergency lighting power supply management method based on water pressure prediction according to claim 9, characterized in that: In step S4, the dynamic adjustment includes: When it is predicted that the water pressure will be higher than the first preset threshold within the next T time period, the hydropower generation unit will be controlled to generate electricity at the rated power, and the surplus electrical energy will be stored in the energy storage management unit and allocated to non-critical loads. When it is predicted that the water pressure will be between the first and second preset thresholds within the next T time period, the hydropower generation unit will be controlled to operate at the optimal efficiency point, and the generated electricity will be prioritized for energy storage charging. When it is predicted that the water pressure will be lower than the second preset threshold within a future time period T, the hydropower generation unit will reduce its power generation or suspend power generation, and the energy storage management unit will maintain the power supply for emergency lighting loads.

Images