Secondary water supply pump set operation optimization method based on minimum energy consumption

Abstract

The application discloses a secondary water supply pump group operation optimization method based on minimum energy consumption, relates to the technical field of water supply equipment, and comprises a three-layer intelligent decision-making scheduling architecture, the three-layer intelligent decision-making scheduling architecture comprising a real-time control layer, an optimization calculation layer and a strategy scheduling layer; generating an actual efficiency cloud picture, sending the real-time efficiency cloud picture to the optimization calculation layer; constructing a water consumption database, constructing a water consumption mode library, generating a predicted water consumption curve, and sending the predicted water consumption curve to the optimization calculation layer; obtaining representative partition points of a pipe network, running a test, obtaining a pipe network pressure model, and sending the pipe network pressure model to the optimization calculation layer; the optimization calculation layer obtains current electricity price information, and formulates a pump group combination scheme in combination with the real-time efficiency cloud picture, the predicted water consumption curve and the pipe network pressure model; a super-low-power pressure stabilizing pump and a small air pressure tank are additionally arranged in the pump group, and the pump group enters a hibernation mode. The application has the effects of improving the energy consumption efficiency and intelligence of the secondary water supply pump group and reducing the energy consumption level of the pump group.

Description

Technical Field

[0001] This application relates to the technical field of water supply equipment, and in particular to a method for optimizing the operation of secondary water supply pump sets based on the lowest energy consumption. Background Technology

[0002] Secondary water supply pumping stations are a crucial component of urban public water supply systems. They are primarily used to pressurize tap water a second time and deliver it to high-rise buildings or homes of users located far away.

[0003] In existing technologies, secondary water supply pumping stations consume significant energy for pumping operations. In practice, due to the unpredictable nature of user water consumption and usage times, the pumping equipment in the station needs to be frequently started and operated, which is highly wasteful from an energy consumption perspective. Furthermore, aging pipelines, changes in valve conditions, and alterations in hydraulic characteristics caused by new user connections can lead to inconsistencies in original parameter settings, resulting in substantial temporary energy consumption challenges during later water supply operations. Therefore, optimizing pump unit operation efficiently, intelligently, and while maintaining low energy consumption is a pressing issue that needs to be addressed. Summary of the Invention

[0004] The purpose of this invention is to provide an optimization method for the operation of secondary water supply pump sets with the lowest energy consumption, so as to solve the problems mentioned in the background art.

[0005] This application provides a method for optimizing the operation of a secondary water supply pump set based on the lowest energy consumption, the method comprising: A three-layer intelligent decision-making and scheduling architecture is constructed, which includes a real-time control layer, an optimization calculation layer, and a policy scheduling layer. The policy scheduling layer performs long-term equipment health and policy evaluation and updates it to the optimization calculation layer. Real-time flow data, real-time head data, and real-time shaft power of each operating pump are collected in real time to obtain the current actual efficiency of each operating pump. Based on multiple current actual efficiencies, an actual efficiency cloud map is generated and the real-time efficiency cloud map is sent to the optimization calculation layer. Historical data is acquired to construct a water consumption database, and a water consumption pattern library is constructed based on the water consumption database. Then, the date, weather, and events of the day are acquired. Combining the water consumption database and the water consumption pattern library, a predicted water consumption curve is generated and sent to the optimization calculation layer. Representative partition points of the pipeline network are obtained, and pressure sensors are set up. Multiple operation tests are conducted under different conditions during the nighttime low water consumption period to obtain multiple sets of pump outlet flow and partition point pressure data. The pipeline network pressure model is then obtained and sent to the optimization calculation layer. The optimization calculation layer obtains the current electricity price information and, in conjunction with the real-time efficiency cloud map, predicted water consumption curve, and pipeline pressure model, formulates a pump group combination scheme and sends it to the real-time control layer for strategy implementation. An ultra-low power pressure-stabilizing pump and a small pressure tank are added to the pump set. The optimization calculation layer monitors the total flow rate. When the total flow rate is lower than the preset minimum threshold and stabilizes, the pump set switches to the ultra-low power pressure-stabilizing pump and the small pressure tank to work and enters the sleep mode.

[0006] Preferably, the step of the policy scheduling layer performing long-term device health and policy evaluation and updating it in the optimization calculation layer specifically includes: The strategy scheduling layer obtains long-term operating data of the secondary water supply pump group, extracts the cumulative operating time of each device in the pump group, and evaluates the health of each device to obtain a health reference value. Based on the cumulative running time and the health reference value, add a running priority weight to each device in the pump group, and generate and update the rotation order table of each device in the pump group according to the running priority weight; The long-term operating efficiency of each device is obtained based on the cumulative operating time and the health reference value; Determine whether the long-term operating efficiency is consistently lower than a preset standard threshold for a preset time period; If it is determined that the long-term operating efficiency is consistently lower than the standard threshold, then the operating priority weight of the device is identified and marked as the target weight, and the target weight is reduced.

[0007] Preferably, the steps of collecting real-time flow data, real-time head data, and real-time shaft power of each operating pump in real time to obtain the current actual efficiency of each operating pump, and generating an actual efficiency cloud map based on multiple current actual efficiencies, are as follows: Real-time flow data, real-time head data, and real-time shaft power of each operating pump are collected in real time. Based on the real-time flow data and the real-time head data, the actual energy consumption parameters are obtained. Based on the real-time shaft power and the actual energy consumption parameters, the current actual efficiency of each operating pump under the current operating conditions is calculated. Obtain the fixed pump performance curve for each operating pump, generate a pump performance curve library, and generate a theoretical efficiency cloud map of the pump set at the time of manufacture based on the pump performance curve library. The theoretical efficiency cloud map is updated based on the current actual efficiency to obtain the actual efficiency cloud map. The current actual efficiency is monitored, and the actual efficiency cloud map is adjusted when the current actual efficiency changes.

[0008] Preferably, the steps of acquiring historical data, constructing a water consumption database, constructing a water consumption pattern library based on the water consumption database, acquiring the current date, current weather, and current events, and combining the water consumption database and the water consumption pattern library to generate a predicted water consumption curve are as follows: Acquire historical data, including historical time points, historical weather, historical special events, and historical water consumption, and construct a water consumption database; Based on the historical time points, historical weather, and historical special events, the historical water consumption is classified to construct a water consumption pattern database; Before the start of each day, obtain the date, weather, and events of the day, and generate a daily information package; Based on the information package of the day, match it in the water use pattern database to obtain multiple similar historical days, and extract the historical water use curve of each historical day in the water consumption database; The similarity of each historical day is extracted, and the historical water use curves are weighted according to the similarity and then fused to obtain the predicted water use curve for that day.

[0009] Preferably, the steps for obtaining representative sections of the pipeline network and setting up pressure sensors to conduct operational tests under multiple different conditions during off-peak water usage periods at night, obtaining multiple sets of pump outlet flow and section pressure data, and further obtaining the pipeline network pressure model are as follows: Obtain network information of the pipeline network, identify water flow conflict areas in the network information, obtain representative partition points that are unfavorable to water flow, and set pressure sensors at the location of each representative partition point; During off-peak water usage periods at night, start one pump at multiple different and stable speeds and record the pump outlet flow rate and zone pressure data corresponding to each speed. Based on the pump outlet flow rate and the zone pressure data, the static head parameter and pipe friction coefficient in the pipeline are obtained. Based on the static head parameter and the pipe friction coefficient, the water flow pressure in the pipe network is modeled to obtain the pipe network pressure model.

[0010] Preferably, the optimization calculation layer obtains the current electricity price information, and, in conjunction with the real-time efficiency cloud map, predicted water consumption curve, and pipeline pressure model, formulates a pump group combination scheme, and sends it to the real-time control layer for strategy implementation. Specifically, the steps are as follows: The optimization calculation layer obtains theoretical energy consumption information based on the real-time efficiency cloud map and the predicted water consumption curve; Based on the pipeline pressure model, information on the additional energy consumption required by the pump unit during operation is obtained; The optimization calculation layer obtains the current electricity price information, combines it with the theoretical energy consumption information and the additional energy consumption information, and obtains the actual energy consumption data. Based on the current electricity price information and the actual energy consumption data, the pump group equipment arrangement and combination query is performed to obtain the pump group combination scheme with the lowest cost, and the pump group combination scheme is sent to the actual control layer for strategy implementation.

[0011] Preferably, after formulating the pump group combination scheme and issuing it to the real-time control layer for strategy implementation, the method further includes: During operation, the optimization calculation layer collects actual water consumption data according to a preset time period to obtain an actual water consumption curve. The curve change of the actual water consumption curve is identified to obtain the direction of the curve change of the actual water consumption, and the trend of the actual water consumption change is determined based on the direction of the curve change. Extract the predicted water consumption from the predicted water consumption curve within the preset time period, compare the actual water consumption with the predicted water consumption, and obtain the prediction deviation value. Based on the predicted deviation value and the changing trend, the predicted water consumption curve is rolled over and corrected within a preset small range.

[0012] Preferably, the optimization calculation layer monitors the total flow rate. When the total flow rate is lower than a preset minimum threshold and stabilizes, the pump group switches to ultra-low power pressure stabilizing pump and small pressure tank for operation, entering a sleep mode. The specific steps are as follows: The optimization computing layer monitors the total traffic and determines whether the total traffic is lower than a preset minimum threshold. If the total flow rate is determined to be lower than the minimum threshold and the flow rate remains stable within a preset time period, the pump set switches to ultra-low power pressure stabilizing pump and small pressure tank for operation, and the pump set enters sleep mode. The hibernation modes include a maintenance mode and a compensation mode; The maintenance mode is specifically as follows: All main pumps in the pump set are shut down, and only the ultra-low power pressure stabilizing pump is used to maintain the basic pressure in the pipeline network according to the static head parameters. The compensation mode is specifically as follows: All main pumps in the pump set are shut down, and water is supplied to the pipeline network and pressure is maintained solely by small pressure tanks. Pressure changes in the small pressure tanks are monitored in real time. Once the pressure in the small pressure tank drops to the preset lower limit, the small pressure tank is pressurized by an ultra-low power pressure stabilizing pump and enters the next pressure stabilization cycle. Among these steps, the fluctuation of the total flow rate is identified to obtain the fluctuation parameter value of the total flow rate; When the fluctuation parameter value is greater than the preset standard value, the sleep mode is the maintenance mode; when the fluctuation parameter is equal to or less than the preset standard value, the maintenance mode is switched to the compensation mode.

[0013] In summary, this application includes at least one of the following beneficial technical effects: By constructing a three-layer intelligent decision-making and scheduling architecture, including a real-time control layer, an optimization calculation layer, and a strategy scheduling layer, the operating status of the pump sets is optimized. Real-time flow data, real-time head data, and real-time shaft power data of each operating pump are collected to obtain the current actual efficiency of each pump. An actual efficiency cloud map is then generated and sent to the optimization calculation layer. Next, historical data is acquired to construct a water consumption database and a water consumption pattern library, generating a predicted water consumption curve, which is also sent to the optimization calculation layer. Then, representative zoning points in the pipeline network are acquired, and pressure sensors are installed to conduct tests within the network, obtaining multiple sets of pump outlet flow and zoning point pressure data to obtain the pipeline network pressure pattern, which is sent to the optimization calculation layer. The optimization calculation layer then acquires current electricity price information, combines it with the real-time efficiency cloud map, predicted water consumption curve, and pipeline network pressure model to formulate a pump set combination scheme, which is then sent to the real-time control layer for strategy implementation. Finally, an ultra-low power stabilizing pump and a small pressure tank are added to the pump set. The total flow rate in the pipeline network is monitored. When the total flow rate is extremely low and stable, the pump set enters a sleep mode and switches to low-power pressure-stabilizing pumps and small pressure tanks for operation. This improves the energy efficiency and intelligence of the secondary water supply pump set and reduces the energy consumption level of the pump set. Attached Figure Description

[0014] Figure 1 This is the step flow of the method for optimizing the operation of a secondary water supply pump set based on the lowest energy consumption, provided in the embodiments of this application. Detailed Implementation

[0015] The following combination Figure 1 This application will be described in further detail, but the embodiments of the present invention are not limited thereto. This application discloses an optimization method for the operation of a secondary water supply pump set based on the lowest energy consumption.

[0016] In this embodiment, the method for optimizing the operation of a secondary water supply pump set with the lowest energy consumption includes: S100: Construct a three-layer intelligent decision-making and scheduling architecture, which includes a real-time control layer, an optimization computing layer, and a policy scheduling layer. The policy scheduling layer performs long-term device health and policy evaluation and updates it to the optimization computing layer. S200: Real-time acquisition of real-time flow data, real-time head data and real-time shaft power of each operating pump to obtain the current actual efficiency of each operating pump, and generate an actual efficiency cloud map based on multiple current actual efficiencies, and send the real-time efficiency cloud map to the optimization calculation layer. S300: Acquire historical data, build a water consumption database, build a water consumption pattern library based on the water consumption database, acquire the current date, current weather and current events, combine the water consumption database and water consumption pattern library, generate a predicted water consumption curve, and send it to the optimization calculation layer. S400: Obtain representative partition points of the pipeline network and set up pressure sensors to conduct multiple operation tests under different conditions during the nighttime low water consumption period, obtain multiple sets of pump outlet flow and partition point pressure data, further obtain the pipeline network pressure model, and send it to the optimization calculation layer; S500: The optimization calculation layer obtains the current electricity price information, and combines it with the real-time efficiency cloud map, predicted water consumption curve and pipeline pressure model to formulate pump group combination schemes and send them to the real-time control layer for strategy implementation; S600: An ultra-low power pressure-stabilizing pump and a small pressure tank are added to the pump set. The optimized calculation layer monitors the total flow rate. When the total flow rate is lower than the preset minimum threshold and stabilizes, the pump set switches to the ultra-low power pressure-stabilizing pump and the small pressure tank to work and enters the sleep mode.

[0017] The policy scheduling layer performs long-term device health and policy evaluations and updates the steps in the optimization calculation layer, specifically as follows: The strategy scheduling layer obtains long-term operating data of the secondary water supply pump set, extracts the cumulative operating time of each device in the pump set, and evaluates the health of each device to obtain a health reference value. Based on the cumulative running time and health reference value, add running priority weight to each device in the pump group, and generate and update the rotation order table of each device in the pump group according to the running priority weight; The long-term operating efficiency of each device is obtained based on the cumulative operating time and health reference values; Determine whether the long-term operating efficiency is consistently lower than a preset standard threshold within a preset time period; If it is determined that the long-term operating efficiency is consistently below the standard threshold, the operating priority weight of the device is identified and marked as the target weight, and the target weight is reduced.

[0018] In practice, taking the secondary water supply pump room of a 30-story residential building in a certain city as an example, the pump room is equipped with three main pumps (numbered A, B, and C). The strategy scheduling layer first collects the operating data of the past year and extracts the cumulative operating time of each pump. For example, pump A has operated for 4,500 hours, pump B for 3,800 hours, and pump C for 5,000 hours. Then, the system evaluates the health of each pump based on vibration monitoring data and historical maintenance records. The health reference value for pump A is 92 (out of 100), for pump B it is 88, and for pump C it is 85. Based on the cumulative operating time and health reference value, the system adds an operating priority weight to each pump. Because pump C has the longest operating time and a relatively low health, its weight is set to a lower value of 1; pump A has a moderate operating time and the highest health, so its weight is set to the highest value of 3; and pump B's weight is set to 2. According to these weights, the system generates and updates the rotation order table, prioritizing the scheduling of pumps A and B with higher weights for daily water supply, while pump C is used as a standby or for low-load periods. Next, the system calculates the long-term operating efficiency of each pump; for example, pump A has an efficiency of 78%, pump B 75%, and pump C 72%. The system's preset standard efficiency threshold is 74%. The system determines that pump C's efficiency has been below 74% for 30 consecutive days. Therefore, the system identifies pump C's operating priority weight (currently 1), marks it as the target weight, and further reduces this weight to 0.5. This will reduce the frequency of pump C's operation in subsequent scheduling, preventing increased energy consumption due to low efficiency, and simultaneously reminding maintenance personnel to inspect pump C.

[0019] The steps for collecting real-time flow data, real-time head data, and real-time shaft power of each operating pump to obtain the current actual efficiency of each operating pump, and generating an actual efficiency cloud map based on multiple current actual efficiencies, are as follows: Real-time flow data, real-time head data, and real-time shaft power of each operating pump are collected in real time. Based on the real-time flow data and real-time head data, the actual energy consumption parameters are obtained. Based on the real-time shaft power and actual energy consumption parameters, the current actual efficiency of each operating pump under the current operating conditions is calculated. Obtain the fixed pump performance curve for each operating pump, generate a pump performance curve library, and generate a theoretical efficiency cloud map of the pump set at the time of delivery based on the pump performance curve library. The theoretical efficiency cloud map is updated based on the current actual efficiency to obtain the actual efficiency cloud map. The current actual efficiency is monitored, and the actual efficiency cloud map is adjusted when the current actual efficiency changes.

[0020] In practice, taking the secondary water supply pump room of a 30-story residential building in a certain city as an example, the system collects real-time data from each operating pump. For instance, during the peak water usage period at 10:00 AM, pump A's real-time flow rate is 50 cubic meters per hour, its real-time head is 60 meters, and its real-time shaft power is 12 kilowatts. Based on the flow rate and head data, the system calculates the actual energy consumption parameters of pump A under this operating condition. Then, combined with the shaft power of 12 kilowatts, the system calculates that pump A's current actual efficiency under the current operating condition is 75%. Simultaneously, the system retrieves pump A's theoretical performance curve from a pre-stored pump performance curve library. This curve shows that at a flow rate of 50 cubic meters per hour and a head of 60 meters, the theoretical efficiency should be 80%. The system updates the theoretical efficiency cloud map based on the current actual efficiency of 75%, generating an actual efficiency cloud map reflecting pump A's current true performance. The system continuously monitors efficiency. At 3 PM, water consumption decreased, and pump A's flow rate became 30 cubic meters per hour, its head became 55 meters, and its shaft power became 8 kilowatts, resulting in a calculated actual efficiency of 70%. The system immediately detected this change and dynamically adjusted the actual efficiency cloud map, updating pump A's efficiency point under low flow conditions to 70%. In this way, the optimization calculation layer always has the latest and most accurate efficiency distribution of each pump, providing a precise basis for developing energy-saving operation plans.

[0021] The steps for acquiring historical data, constructing a water consumption database, building a water consumption pattern library based on the database, acquiring the current date, weather, and events, and combining the water consumption database and the water consumption pattern library to generate a predicted water consumption curve are as follows: Acquire historical data, including historical time points, historical weather, historical special events, and historical water consumption, and build a water consumption database; Based on historical time points, historical weather, and historical special events, historical water consumption is classified and a water consumption pattern database is constructed. Before the start of each day, obtain the date, weather, and events of the day, and generate a daily information package; Based on the information package of the day, match it in the water use pattern database to obtain multiple similar historical days, and extract the historical water use curve of each historical day from the water consumption database. The similarity of each historical day is extracted, and the historical water use curves are weighted according to the similarity and then fused to obtain the predicted water use curve for the current day.

[0022] In practice, taking the secondary water supply pump room of a 30-story residential building in a certain city as an example, the system of this building's pump room acquired historical data from the past two years to construct a water consumption database. This data includes historical water consumption at each point in time every 24 hours, corresponding historical weather (e.g., sunny, rainy, temperature), and historical special events (e.g., holidays, community activities). Based on this information, the system categorizes water consumption and constructs a water consumption pattern database, such as "weekday sunny pattern," "weekend rainy pattern," and "National Day holiday pattern." Today is June 15th, Wednesday. The system retrieved the information for the day in the early morning: date is June 15th, weather forecast is sunny, high temperature is 35 degrees Celsius, and there are no special community events on that day. The system matches this information package with the water consumption pattern database and finds several similar historical days, such as June 10th (Wednesday, sunny) and June 17th (Friday, sunny) last year. The system extracts the historical water consumption curves for these historical days from the database and calculates the similarity between each historical day and today (based on date proximity, weather matching, etc.). Assuming the similarity is 0.9 for June 10th and 0.7 for June 17th, the system assigns weights based on the similarity, fusing the water consumption curve for June 10th (weight 0.9) with the water consumption curve for June 17th (weight 0.7). For example, at the morning peak of 8:00 AM, the historical water consumption on June 10th was 40 tons, and on June 17th it was 45 tons. After fusion, the predicted water consumption for that day is (400.9 + 450.7) / (0.9 + 0.7) ≈ 42.2 tons. By performing this type of fusion calculation at various time points throughout the day, a smooth predicted water consumption curve reflecting the expected water consumption on June 15th is finally generated and sent to the optimization calculation layer.

[0023] The steps to obtain representative sections of the pipeline network and install pressure sensors, and conduct operational tests under various conditions during off-peak water usage periods at night, to obtain multiple sets of pump outlet flow and section pressure data, and further derive the pipeline network pressure model, are as follows: Obtain network information of the pipeline network, identify water flow conflict areas in the network information, obtain representative partition points that are unfavorable to water flow, and set pressure sensors at the location of each representative partition point. During off-peak water usage periods at night, start one pump at multiple different and stable speeds and record the pump outlet flow rate and zone pressure data corresponding to each speed. Based on the pump outlet flow rate and zone pressure data, the static head parameter and pipe friction coefficient in the pipeline are obtained. Based on the static head parameter and the pipe friction coefficient, the water flow pressure in the pipe network is modeled to obtain the pipe network pressure model.

[0024] In this application, taking the secondary water supply pump room of a 30-story residential building in a certain city as an example, in order to establish an accurate pipeline pressure model, the system first obtained a complete network information map of the building's water supply network. By analyzing the water flow path and pipe diameter changes, the system identified several representative partition points that are unfavorable to water flow and may cause significant resistance losses, such as the pipe bends on the 15th and 25th floors, and the water outlet at the farthest end of the roof. High-precision pressure sensors were installed at these locations. During the nighttime low water consumption period from 2:00 AM to 4:00 AM, the system started pump A for testing. Pump A operated at three stable speeds of 800 rpm, 1000 rpm, and 1200 rpm, with each run lasting 15 minutes. The system recorded the corresponding pump outlet flow rate (e.g., 20, 30, and 40 cubic meters per hour) at each speed, as well as the pressure data fed back by the pressure sensors at each partition point (e.g., pressure on the 15th floor was 0.32, 0.38, and 0.45 MPa). Based on these paired flow-pressure data, the system uses a mathematical model to calculate the static head parameter (i.e., the base pressure required to overcome gravity and achieve the desired height) of the pipeline system as 0.25 MPa, as well as the comprehensive friction coefficient of the pipes. Based on the calculated static head parameter and pipe friction coefficient, the system successfully constructed a pipeline pressure model for the residential building. This model can accurately predict the pressure values ​​that need to be maintained at key points in the pipeline network under any given flow rate, providing a crucial foundation for optimizing pump head settings and avoiding energy waste.

[0025] The optimization calculation layer obtains current electricity price information and, in conjunction with real-time efficiency cloud maps, predicted water consumption curves, and pipeline pressure models, formulates pump unit combination schemes and distributes them to the real-time control layer for strategy implementation. The specific steps are as follows: The optimization computing layer obtains theoretical energy consumption information based on real-time efficiency cloud maps and predicted water consumption curves; Based on the pipeline pressure model, information on the additional energy consumption required by the pump unit during operation is obtained; The optimization calculation layer obtains the current electricity price information, combines it with theoretical energy consumption information and additional energy consumption information to obtain the actual energy consumption data; Based on current electricity price information and actual energy consumption data, the pump unit equipment arrangement and combination query is performed to obtain the pump unit combination scheme with the lowest cost, and the pump unit combination scheme is sent to the actual control layer for strategy implementation.

[0026] In practice, taking the secondary water supply pump room of a 30-story residential building in a certain city as an example, the optimization calculation layer first obtains the current peak and off-peak electricity price information when formulating the daily pump operation plan: the peak hour (8:00-22:00) electricity price is 1.2 yuan / kWh, and the off-peak hour (22:00-8:00 the next day) electricity price is 0.6 yuan / kWh. Then, it combines three aspects of information for calculation. The first is the real-time efficiency cloud map, which shows that pump A has the highest efficiency (78%) under medium flow, pump B has good efficiency (76%) under low flow, and pump C has generally low efficiency. The second is the predicted water consumption curve, which shows that the water demand is the largest during the morning peak (8:00-10:00), about 50 tons / hour. The third is the pipeline pressure model, which indicates that in order to meet the water supply pressure at the most unfavorable point under this flow, the pump set needs to provide an outlet pressure of at least 0.5 MPa. Based on the predicted water consumption curve and the pipeline pressure model, the optimization calculation layer calculates the theoretical energy consumption information to meet the water supply demand. Next, it combines real-time efficiency cloud maps to estimate the additional energy consumption generated by operating a single pump or a combination of different pumps. For example, calculations show that during the morning peak hours, using the most efficient pump A alone consumes 15 kilowatts of energy; while using a combination of pumps A and B, although the total flow rate is sufficient, the actual energy consumption may reach 16.5 kilowatts because pump B is not optimally efficient during this period. The optimization calculation layer then calculates the electricity cost of different schemes based on the current peak electricity price. After querying and calculating all possible pump combinations, it arrives at the lowest-cost scheme: using pump A alone during the morning peak hours, using pump B alone during the midday low-load period, and using pump C for a period of time during the late-night off-peak hours to balance equipment wear. This optimal pump combination scheme is immediately sent to the real-time control layer for execution.

[0027] After formulating the pump unit combination plan and issuing it to the real-time control layer for strategy implementation, the following steps are also included: During operation, the optimization computing layer collects actual water consumption data according to a preset time period to obtain an actual water consumption curve. The curve change of the actual water consumption curve is identified to obtain the direction of the curve change of the actual water consumption, and the trend of the actual water consumption change is determined based on the direction of the curve change. Extract the predicted water consumption curve within a preset time period, compare the actual water consumption with the actual water consumption, and obtain the prediction deviation value. Based on the predicted deviation and its trend, the predicted water consumption curve is rolled over and corrected within a preset small range.

[0028] In practice, taking the secondary water supply pump room of a 30-story residential building in a certain city as an example, after the optimized calculation layer issued the pump group combination scheme and started operation, the system did not stop optimizing. It continuously collected actual water consumption data from the main water supply meter every hour, according to a preset time period. For example, during the period from 9:00 AM to 10:00 AM, the predicted water consumption curve showed a consumption of 48 tons, but the actual collected water consumption curve showed that due to a sudden rise in temperature that morning, residential water consumption increased, reaching 52 tons. The system identified the actual water consumption curve for this period and found that the curve showed a rapid upward trend starting at 9:30 AM, thus confirming that the actual water consumption was changing upwards. The system extracted the predicted value of 48 tons from the predicted water consumption curve for this period (9:00-10:00 AM) and compared it with the actual water consumption of 52 tons, obtaining a prediction deviation value of +4 tons (i.e., 4 tons were underestimated). The system doesn't immediately and drastically modify the daily forecast curve. Instead, based on the deviation (+4 tons) and the upward trend, it makes rolling corrections for the next short period (the forecast water consumption curve for the following hour). For example, it might slightly revise the original forecast water consumption from 45 tons to 48 tons between 10:00 and 11:00. This rolling correction allows the forecast water consumption curve to dynamically adjust in line with actual water consumption, enabling the optimization calculation layer to fine-tune pump operation strategies based on more realistic water consumption forecasts, continuously approaching the minimum energy consumption target.

[0029] The optimization calculation layer monitors the total flow rate. When the total flow rate falls below a preset minimum threshold and stabilizes, the pump unit switches to ultra-low power pressure-stabilizing pump and small pressure tank to operate, entering sleep mode. The specific steps are as follows: The optimization computing layer monitors the total traffic and determines whether the total traffic is below a preset minimum threshold. If the total flow rate is determined to be below the minimum threshold and remains stable within a preset time period, the pump unit will switch to ultra-low power pressure stabilizing pump and small pressure tank for operation, and the pump unit will enter sleep mode. Hibernation modes include sustain mode and compensation mode; The maintenance mode is as follows: All the main pumps in the pump set are shut down, and only the ultra-low power pressure stabilizing pump is used to maintain the basic pressure in the pipeline according to the static head parameter. The specific compensation model is as follows: All main pumps in the pump set are shut down, and water is supplied to the pipeline network and pressure is maintained solely by small pressure tanks. Pressure changes in the small pressure tanks are monitored in real time. Once the pressure in the small pressure tank drops to the preset lower limit, the small pressure tank is pressurized by an ultra-low power pressure stabilizing pump and enters the next pressure stabilization cycle. Among these steps, the fluctuations in total flow are identified, and the fluctuation parameter values ​​of total flow are obtained. When the fluctuation parameter value is greater than the preset standard value, the sleep mode is the maintenance mode. When the fluctuation parameter is equal to or less than the preset standard value, the maintenance mode is switched to the compensation mode.

[0030] In practice, taking the secondary water supply pump room of a 30-story residential building in a certain city as an example, the pump room added an ultra-low power stabilizing pump with a power of only 0.75 kW and a small pressure tank to the pump set. The optimized calculation layer continuously monitors the total outflow of the pipeline network. At 3:00 AM, the system found that the total flow rate was consistently lower than the preset minimum threshold (e.g., 2 cubic meters / hour), and this low flow rate state had been stable for more than 20 minutes. The system determined that the conditions for entering the sleep mode were met. First, the system identified the fluctuation of the total flow rate data and calculated the standard deviation of the flow rate over the past 10 minutes as the fluctuation parameter value. Assuming that there was only sporadic water use during this period, the flow rate fluctuation was very small, and the fluctuation parameter value was calculated to be 0.3, which is less than the preset standard value of 1.0. Therefore, the system decided to enter the sleep mode using the compensation mode. The specific operation is as follows: the system smoothly shuts down all running pumps A and B. Then, it relies solely on the water and air pressure stored in the small pressure tank to replenish water to the pipeline network to maintain the basic water pressure at the user end. The system monitors the pressure inside the pressure tank in real time. When the pressure gradually decreases due to water supply, from 0.4 MPa to the preset lower limit of 0.25 MPa, the ultra-low power pressure stabilizing pump automatically starts to quickly pressurize the pressure tank, raising the pressure back to 0.4 MPa before stopping, completing one pressure stabilization cycle. During this process, the main pump completely stops, consuming only a very small amount of electricity for the intermittent short-term operation of the pressure stabilizing pump and the energy storage of the pressure tank, thus achieving a deep energy-saving dormant mode during periods of extremely low water consumption. If the fluctuation parameter value is greater than 1.0, the system will adopt a maintenance mode, that is, the ultra-low power pressure stabilizing pump will operate at low speed to maintain the basic pressure of the pipeline network.

[0031] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A method for optimizing the operation of a secondary water supply pump set with the lowest energy consumption, characterized in that: include: A three-layer intelligent decision-making and scheduling architecture is constructed, which includes a real-time control layer, an optimization calculation layer, and a policy scheduling layer. The policy scheduling layer performs long-term equipment health and policy evaluation and updates it to the optimization calculation layer. Real-time flow data, real-time head data, and real-time shaft power of each operating pump are collected in real time to obtain the current actual efficiency of each operating pump. Based on multiple current actual efficiencies, an actual efficiency cloud map is generated and the real-time efficiency cloud map is sent to the optimization calculation layer. Historical data is acquired to construct a water consumption database, and a water consumption pattern library is constructed based on the water consumption database. Then, the date, weather, and events of the day are acquired. Combining the water consumption database and the water consumption pattern library, a predicted water consumption curve is generated and sent to the optimization calculation layer. Representative partition points of the pipeline network are obtained, and pressure sensors are set up. Multiple operation tests are conducted under different conditions during the nighttime low water consumption period to obtain multiple sets of pump outlet flow and partition point pressure data. The pipeline network pressure model is then obtained and sent to the optimization calculation layer. The optimization calculation layer obtains the current electricity price information and, in conjunction with the real-time efficiency cloud map, predicted water consumption curve, and pipeline pressure model, formulates a pump group combination scheme and sends it to the real-time control layer for strategy implementation. An ultra-low power pressure-stabilizing pump and a small pressure tank are added to the pump set. The optimization calculation layer monitors the total flow rate. When the total flow rate is lower than the preset minimum threshold and stabilizes, the pump set switches to the ultra-low power pressure-stabilizing pump and the small pressure tank to work and enters the sleep mode.

2. The method for optimizing the operation of a secondary water supply pump set based on the lowest energy consumption, as described in claim 1, is characterized in that... The steps of the policy scheduling layer performing long-term device health and policy evaluations and updating them in the optimization calculation layer are as follows: The strategy scheduling layer obtains long-term operating data of the secondary water supply pump group, extracts the cumulative operating time of each device in the pump group, and evaluates the health of each device to obtain a health reference value. Based on the cumulative running time and the health reference value, add a running priority weight to each device in the pump group, and generate and update the rotation order table of each device in the pump group according to the running priority weight; The long-term operating efficiency of each device is obtained based on the cumulative operating time and the health reference value; Determine whether the long-term operating efficiency is consistently lower than a preset standard threshold for a preset time period; If it is determined that the long-term operating efficiency is consistently lower than the standard threshold, then the operating priority weight of the device is identified and marked as the target weight, and the target weight is reduced.

3. The method for optimizing the operation of a secondary water supply pump set based on the lowest energy consumption, as described in claim 2, is characterized in that... The steps of collecting real-time flow data, real-time head data, and real-time shaft power of each operating pump in real time to obtain the current actual efficiency of each operating pump, and generating an actual efficiency cloud map based on multiple current actual efficiencies, are as follows: Real-time flow data, real-time head data, and real-time shaft power of each operating pump are collected in real time. Based on the real-time flow data and the real-time head data, the actual energy consumption parameters are obtained. Based on the real-time shaft power and the actual energy consumption parameters, the current actual efficiency of each operating pump under the current operating conditions is calculated. Obtain the fixed pump performance curve for each operating pump, generate a pump performance curve library, and generate a theoretical efficiency cloud map of the pump set at the time of manufacture based on the pump performance curve library. The theoretical efficiency cloud map is updated based on the current actual efficiency to obtain the actual efficiency cloud map. The current actual efficiency is monitored, and the actual efficiency cloud map is adjusted when the current actual efficiency changes.

4. The method for optimizing the operation of a secondary water supply pump set based on the lowest energy consumption, as described in claim 3, is characterized in that... The steps of acquiring historical data, constructing a water consumption database, constructing a water consumption pattern library based on the water consumption database, acquiring the current date, current weather, and current events, and combining the water consumption database and the water consumption pattern library to generate a predicted water consumption curve are as follows: Acquire historical data, including historical time points, historical weather, historical special events, and historical water consumption, and construct a water consumption database; Based on the historical time points, historical weather, and historical special events, the historical water consumption is classified to construct a water consumption pattern database; Before the start of each day, obtain the date, weather, and events of the day, and generate a daily information package; Based on the information package of the day, match it in the water use pattern database to obtain multiple similar historical days, and extract the historical water use curve of each historical day in the water consumption database; The similarity of each historical day is extracted, and the historical water use curves are weighted according to the similarity and then fused to obtain the predicted water use curve for that day.

5. The method for optimizing the operation of a secondary water supply pump set based on the lowest energy consumption, as described in claim 4, is characterized in that... The steps to obtain representative sections of the pipeline network and install pressure sensors, and conduct operational tests under various conditions during off-peak water usage periods at night, to obtain multiple sets of pump outlet flow and section pressure data, and further derive the pipeline network pressure model, are as follows: Obtain network information of the pipeline network, identify water flow conflict areas in the network information, obtain representative partition points that are unfavorable to water flow, and set pressure sensors at the location of each representative partition point; During off-peak water usage periods at night, start one pump at multiple different and stable speeds and record the pump outlet flow rate and zone pressure data corresponding to each speed. Based on the pump outlet flow rate and the zone pressure data, the static head parameter and pipe friction coefficient in the pipeline are obtained. Based on the static head parameter and the pipe friction coefficient, the water flow pressure in the pipe network is modeled to obtain the pipe network pressure model.

6. The method for optimizing the operation of a secondary water supply pump set based on the lowest energy consumption, as described in claim 5, is characterized in that... The optimization calculation layer obtains the current electricity price information and, in conjunction with the real-time efficiency cloud map, predicted water consumption curve, and pipeline pressure model, formulates a pump group combination scheme and sends it to the real-time control layer for strategy implementation. The specific steps are as follows: The optimization calculation layer obtains theoretical energy consumption information based on the real-time efficiency cloud map and the predicted water consumption curve; Based on the pipeline pressure model, information on the additional energy consumption required by the pump unit during operation is obtained; The optimization calculation layer obtains the current electricity price information, combines it with the theoretical energy consumption information and the additional energy consumption information, and obtains the actual energy consumption data. Based on the current electricity price information and the actual energy consumption data, the pump group equipment arrangement and combination query is performed to obtain the pump group combination scheme with the lowest cost, and the pump group combination scheme is sent to the actual control layer for strategy implementation.

7. The method for optimizing the operation of a secondary water supply pump set based on the lowest energy consumption, as described in claim 6, is characterized in that... After formulating the pump unit combination scheme and issuing it to the real-time control layer for strategy implementation, the process also includes: During operation, the optimization calculation layer collects actual water consumption data according to a preset time period to obtain an actual water consumption curve. The curve change of the actual water consumption curve is identified to obtain the direction of the curve change of the actual water consumption, and the trend of the actual water consumption change is determined based on the direction of the curve change. Extract the predicted water consumption from the predicted water consumption curve within the preset time period, compare the actual water consumption with the predicted water consumption, and obtain the prediction deviation value. Based on the predicted deviation value and the changing trend, the predicted water consumption curve is rolled over and corrected within a preset small range.

8. The method for optimizing the operation of a secondary water supply pump set based on the lowest energy consumption, as described in claim 7, is characterized in that... The optimization calculation layer monitors the total flow rate. When the total flow rate falls below a preset minimum threshold and stabilizes, the pump unit switches to ultra-low power pressure-stabilizing pump and small pressure tank for operation, entering a sleep mode. The specific steps are as follows: The optimization computing layer monitors the total traffic and determines whether the total traffic is lower than a preset minimum threshold. If the total flow rate is determined to be lower than the minimum threshold and the flow rate remains stable within a preset time period, the pump set switches to ultra-low power pressure stabilizing pump and small pressure tank for operation, and the pump set enters sleep mode. The hibernation modes include a maintenance mode and a compensation mode; The maintenance mode is specifically as follows: All main pumps in the pump set are shut down, and only the ultra-low power pressure stabilizing pump is used to maintain the basic pressure in the pipeline network according to the static head parameters. The compensation mode is specifically as follows: All main pumps in the pump set are shut down, and water is supplied to the pipeline network and pressure is maintained solely by small pressure tanks. Pressure changes in the small pressure tanks are monitored in real time. Once the pressure in the small pressure tank drops to the preset lower limit, the small pressure tank is pressurized by an ultra-low power pressure stabilizing pump and enters the next pressure stabilization cycle. Among these steps, the fluctuation of the total flow rate is identified to obtain the fluctuation parameter value of the total flow rate; When the fluctuation parameter value is greater than the preset standard value, the sleep mode is the maintenance mode; when the fluctuation parameter is equal to or less than the preset standard value, the maintenance mode is switched to the compensation mode.

Images