Thermal balance method for central heating

Abstract

The invention provides a thermal balance method for central heating. The thermal balance method comprises the following steps that a PID controller is used for periodically constructing the compensation weight of each household according to the static parameter and the target adjusting temperature of each household; obtaining a weight normalization parameter matrix of all users; calculating the target traffic of each user according to the weight normalization parameter matrix of all users; the opening degree of a valve of each household is controlled according to the Qtargeti; the problem of unbalanced heat supply can be effectively relieved, and the satisfaction degree of users and the energy-saving effect can be remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of centralized heating technology, and in particular to a method for achieving thermal balance in centralized heating. Background Technology

[0002] In cold northern regions, centralized heating systems are the primary means of winter heating, providing warmth to an entire community or building complex through one or more heat sources. While such systems can effectively improve energy efficiency and reduce heating costs for individual households, they face numerous challenges in practical operation.

[0003] For example, uneven heating can manifest as uneven flow distribution: due to factors such as pipe layout and distance from the heat source, there are often significant differences in flow between different households. Some households closer to the heat source may receive excessive heat, while those farther away may not receive enough. Alternatively, differences in building characteristics can also play a role: the size, orientation, and floor location (e.g., top floor, corner unit) of each residence all affect its heat demand. For instance, larger residences require more heat to maintain a comfortable indoor temperature; corner units or top-floor residences, due to their larger exposed exterior walls, dissipate heat faster and therefore require more heat compensation. Existing solutions include: 1. Manual Adjustment: The traditional solution relies on users adjusting the balancing valve themselves. This method is not only inefficient but also makes precise control difficult. Furthermore, due to a lack of professional knowledge, ordinary users find it difficult to accurately determine the optimal valve opening.

[0004] 2. Simple Automated Control: Some systems attempt to use simple automated control strategies, such as PID controllers based on schedules or fixed parameters, to adjust flow rates. However, these methods often ignore the specific conditions inside the building, making it impossible to meet individual needs.

[0005] 3. Static allocation scheme: Some existing optimization schemes attempt to solve the problem through pre-set traffic allocation rules, but this method fails to take into account the impact of dynamic factors such as weather changes and changes in living patterns, so the effect is limited.

[0006] In recent years, advancements in sensor technology, communication technology, and computing power, particularly the application of the Internet of Things (IoT) and big data analytics, have provided new ideas and technical means to address the aforementioned problems: High-precision sensing equipment: Modern temperature sensors and flow meters can provide more accurate data acquisition, laying a solid foundation for subsequent intelligent decision-making; Real-time data processing and feedback mechanisms: Leveraging cloud computing platforms and edge computing technologies, massive amounts of data can be processed in real time, reacting rapidly and adjusting heating strategies; Personalized control strategies: By combining machine learning algorithms, personalized heating plans can be developed based on each household's specific conditions (such as area and location), thereby improving overall comfort and saving energy. Faced with ever-increasing energy demands and environmental protection pressures, developing an intelligent management system capable of automatically adapting to various complex situations is particularly important. Summary of the Invention

[0007] This invention addresses the problem that existing technologies cannot fully meet the modern society's demand for efficient, precise, and personalized heating. It proposes a centralized heating thermal balance method that can not only effectively alleviate the problem of unbalanced heating but also significantly improve user satisfaction and energy-saving effects.

[0008] To achieve the above objectives, the following technical solution is proposed: A method for achieving heat balance in centralized heating includes the following steps: S1 uses a PID controller to periodically adjust the temperature ΔT based on the static parameters and target values ​​of each household. i Construct the compensation weight W for each household i ; S2, obtain the weight normalization parameter matrix all for all users. weights = [W1, W2, ..., W i ]; S3, according to all weights Calculate the target traffic Qtarget for each household. i ; S4, according to Qtarget i Control the opening degree of the valve for each household; The static parameters include the area parameter S. i / max(S) and position parameters, the position parameters including the edge parameter is corner And the top floor parameter is top ,but: W i =α*S i / max(S)+β*is corner +γ*is top +λ*PID(ΔT i ); Wherein: S i / max(S) represents the area normalized value of the actual heating area of ​​each household and the maximum heating area among all users. The location parameter takes values ​​of 0 or 1. α is the area coefficient, β is the edge coefficient, γ is the top floor coefficient, λ is the temperature sensitivity coefficient, and ΔT is the area coefficient. i Set temperature T for each household set and the current temperature T actual The difference.

[0009] This invention, when calculating the input to the PID controller, takes into account the different needs of various users. This difference can be reflected by adjusting the target temperature or directly modifying the proportional gain of the PID controller. For example, for users with larger areas or in unfavorable locations (such as top floors or corner units), their target temperature can be appropriately increased or the proportional gain can be amplified to provide more heat compensation. This invention not only considers the impact of area size on heat demand but also pays special attention to the impact of location factors on thermal efficiency, achieving more personalized heating management. Based on real-time feedback, the target flow rate is dynamically adjusted: unlike traditional fixed flow rate distribution methods, this system can dynamically adjust the target flow rate for each household based on real-time monitoring data, ensuring optimal temperature balance under different environmental conditions.

[0010] Preferably, step S3 specifically includes the following steps: S31, according to all weights Calculate the total weight and sum weights =sum(all) weights ), for all weights Sum; S32, calculate the target flow for each household: Qtarget i =Q total *(W i / sum weights ), where: Q total This represents the total flow rate of the centralized heating system.

[0011] Preferably, step S4 specifically includes the following steps: S41, Valve opening change: valve opening =(Qtarget) i / Qmax)*100%, where: Qmax is the valve's maximum flow rate; S42 uses a PID controller to adjust the valve based on the valve opening degree. S43, determine whether the current valve has reached its limit position. If yes, the PID controller stops accumulating the integral term and records the over-limit time. S43, determine whether the over-limit time is greater than 30 minutes. If yes, issue an over-limit alarm. If not, return to S42.

[0012] Preferably, step S3 further includes the following steps: S33, determine if the user is an end user; if so, calculate the end user's Qtarget. i Increase the base traffic margin by 5%-10% before outputting the target traffic for each household; otherwise, directly output the target traffic for each household.

[0013] Preferably, step S3 further includes the following step: S34, determine the total target traffic ΣQtarget for all users. i Is it greater than Q? total If so, the target traffic for each household is equal to Qtarget. i *(Q total / ΣQtarget i If not, output the target traffic for each household.

[0014] Preferably, the control cycle of the PDI controller is 15-30 minutes, taking into account thermal inertia; the proportional coefficient Kp, integral coefficient Ki, and derivative coefficient Kd of the PDI controller are obtained using the Ziegler-Nichols empirical method, and the PID (ΔT) i This refers to the situation where, under the set proportional coefficient Kp, integral coefficient Ki, and differential coefficient K, the temperature difference is ΔT. i At that time, the output value of the PDI controller.

[0015] PID Controller Design: Each user's balancing valve heat meter can be considered an independent controlled object. Its input is the target flow rate command issued by the host computer, and its output is the actual flow rate achieved. This invention uses a PID control strategy to dynamically adjust the flow rate for each user, so that the actual temperature of each user is as close as possible to the target temperature.

[0016] As a preferred option, when ΔT i When the value is greater than 0, heating is required. A smoothing factor K is introduced to prevent oscillations, and the flow distribution is increased proportionally. At this time: Qtarget i =Q total *(W i / sum weights )*(1+K*ΔT i ).

[0017] Preferably, step S1 further includes the following step: Considering inertia delay compensation, a Smith predictor is connected to the PID controller. The Smith predictor uses the predicted values ​​to calculate the temperature deviation. ΔT i =Tset -predicted T ; predicted T =T actual +(Q current *heat transfer _ coeff )*delay time ; Among them: predicted T This is the predicted temperature value; Q current Current flow rate; heat transfer _ coeff The heating coefficient; delay time This is a delay time.

[0018] Preferably, step S4 further includes the following step: Considering the impact of temperature changes in adjacent rooms, feedforward control is introduced to compensate for temperature changes between adjacent rooms, as detailed below: neighbor effect =sum((T) neighbor -T set-neighbor *coupling coeff ); feedforward = K f *neighbor effect ; Valve opening degree = valve opening +feedforward; Among them: neighbor effect The parameter representing the influence of temperature on adjacent rooms; T neighbor Temperature of the adjacent room; T set-neighbor Set the temperature for adjacent rooms; coupling coeff The distance correlation coefficient between adjacent rooms; feedforward is the feedforward control parameter; K f This represents the feedforward control coefficient.

[0019] Preferably, S1 further includes the following steps: considering the day-night temperature difference, dynamically adjusting the set temperature, determining whether the current time is the nighttime set value, and if so, adjusting the set temperature to T. set =T set-daytime -T0: Lower the set temperature at night; otherwise, maintain the set temperature at T. set =T set-daytime ; where: T set-daytime Set the temperature value for daytime.

[0020] The beneficial effects of this invention are as follows: When calculating the input to the PID controller, this invention considers the differences in needs among different users and can reflect these differences by adjusting the target temperature or directly modifying the proportional coefficient of the PID controller. For example, for users with larger areas or in unfavorable locations (such as top floors or corner units), their target temperature can be appropriately increased or the proportional coefficient can be amplified to provide more heat compensation. This invention not only considers the impact of area size on heat demand but also pays special attention to the impact of location factors on thermal efficiency, achieving more personalized heating management. Based on real-time feedback, the target flow rate is dynamically adjusted: Unlike traditional fixed flow rate allocation methods, this system can dynamically adjust the target flow rate for each household based on real-time monitoring data, ensuring optimal temperature balance under different environmental conditions. Attached Figure Description

[0021] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0022] Example 1: This embodiment proposes a method for achieving thermal balance in centralized heating, referencing... Figure 1 This includes the following steps: S1 uses a PID controller to periodically adjust the temperature ΔT based on the static parameters and target values ​​of each household. i Construct the compensation weight W for each household i The static parameters include the area parameter S. i / max(S) and position parameters, the position parameters including the edge parameter is corner And the top floor parameter is top ,but: W i =α*S i / max(S)+β*is corner +γ*is top +λ*PID(ΔT i ); Wherein: S i / max(S) represents the area normalized value of the actual heating area of ​​each household and the maximum heating area among all users. The location parameter takes values ​​of 0 or 1. α is the area coefficient, β is the edge coefficient, γ is the top floor coefficient, λ is the temperature sensitivity coefficient, and ΔT is the area coefficient. i Set temperature T for each household set and the current temperature T actual The difference. In this invention, the temperature sensitivity coefficient λ needs to be determined based on the system's thermal inertia experiment. The typical value range of λ is between 0.2 and 0.5. In this embodiment, λ = 0.3 is selected; α, β, and γ are all set empirical values. In this embodiment, α = 0.6, β = 0.3, and γ = 0.2.

[0023] The control cycle of the PDI controller is 15-30 minutes, taking thermal inertia into account; the proportional coefficient Kp, integral coefficient Ki, and derivative coefficient Kd of the PDI controller are obtained using the Ziegler-Nichols empirical method, and the PID (ΔT) i This refers to the situation where, under the set proportional coefficient Kp, integral coefficient Ki, and differential coefficient K, the temperature difference is ΔT. i At that time, the output value of the PDI controller.

[0024] PID Controller Design: Each user's balancing valve heat meter can be considered an independent controlled object. Its input is the target flow rate command issued by the host computer, and its output is the actual flow rate achieved. This invention uses a PID control strategy to dynamically adjust the flow rate for each user, so that the actual temperature of each user is as close as possible to the target temperature.

[0025] The PID controller parameters are set as follows: Proportional term (P): Adjusted directly based on the current error (i.e., the difference between the target temperature and the actual temperature) to quickly respond to temperature changes.

[0026] Integral term (I): Considers the cumulative error over a long period of time, helps to eliminate static error, and is especially suitable for families that are difficult to control precisely through proportional action. It sets the dead zone range to not accumulate integrals and performs anti-saturation treatment of the integral term.

[0027] Differential term (D): Predicts future error trends and prevents overshoot and oscillations, which is especially important when temperatures change rapidly.

[0028] The Ziegler-Nichols empirical method of this invention, referring to Table 1, includes the following specific steps: (1) First set Ki = 0, Kd = 0, and gradually increase Kp until the system exhibits small oscillations; (2) Record Ku and the oscillation period Tu at this time; (3) Calculate the initial values ​​of the PID parameters according to the table above; (4) Fine-tune the parameters until the temperature fluctuation is <±0.3℃; (5) Output Kp, Ki and Kd at this time. In this embodiment, Kp = 0.1, Ki = 0.01, Kd = 0.05; Table 1. Comparison of parameters for the Ziegler-Nichols empirical method. Control type Kp Ti Td P 0.5*Ku - - PI 0.45*Ku 0.83*Tu - PID 0.6*Ku 0.5*Tu 0.12*Tu The following is an example of dynamic adjustment calculation for PID controller parameters: For the parameters of a top-floor user in a certain unit S i =90m 2max(S) = 120m 2 is corner =1, is top =1 T set =22℃, T actual =20℃ to obtain ΔT=T set -T actual = +2℃ Static weight calculation: static_weight=α*Si / max(S)+β*iscorner+γ*istop=0.6*(90 / 120)+0.3*1+0.2*1=0.45+0.3+0.2=0.95 PID controller calculation (Kp = 0.1, Ki = 0.01, Kd = 0.05) pid=PIDController(0.1,0.01,0.05) temp_correction=λ*PID(ΔTi)=0.3*pid(2)=0.25; W i =α*Si / max(S)+β*iscorner+γ*istop+λ*PID(ΔTi)=0.95+0.25=1.2; Personalized PID parameter settings: The proportional, integral, and derivative gains of the PID controller can be customized to suit the specific circumstances of different households, such as room size and location characteristics. This customized approach can more effectively handle various complex situations, improving system response speed and stability.

[0029] Integrated PID control loop design: Each household is equipped with an independent PID controller, which monitors the indoor temperature in real time and dynamically adjusts the opening of the balancing valve to achieve the target flow rate, thereby achieving precise temperature control.

[0030] S2, obtain the weight normalization parameter matrix all for all users. weights = [W1, W2, ..., W i ]; S3, according to all weights Calculate the target traffic Qtarget for each household. i S3 specifically includes the following steps: S31, according to all weights Calculate the total weight and sum weights =sum(all) weights ), for all weights Summation; in this embodiment, all is obtained by substituting the above parameters.weights = [1.2, 0.8, 1.1...]; S32, calculate the target flow for each household: Qtarget i =Q total *(W i / sum weights ), where: Q total =1000m 3 / h represents the total flow rate of the centralized heating system.

[0031] Preferably, when ΔT i When the value is greater than 0, heating is required. A smoothing factor K is introduced to prevent oscillations, and the flow distribution is increased proportionally. At this time: Qtarget i =Q total *(W i / sum weights )*(1+K*ΔT i ).

[0032] S4, according to Qtarget i Controlling the opening degree of each household's valve; S4 specifically includes the following steps: S41, Valve opening change: valve opening =(Qtarget) i / Qmax)*100%, where: Qmax is the valve's maximum flow rate; S42 uses a PID controller to adjust the valve based on the valve opening degree. S43, determine whether the current valve has reached the limit position. If so, the PID controller stops accumulating the integral term, dynamically adjusts the integral limit value to prevent the integral term of the PID controller from being too large, and records the over-limit time. S43, determine if the over-limit time is greater than 30 minutes. If yes, issue an over-limit alarm. If no, return to S42.

[0033] Preferably, considering the influence of temperature changes in adjacent rooms, feedforward control is introduced to compensate for temperature changes in adjacent rooms, as follows: neighbor effect =sum((T) neighbor -T set-neighbor *coupling coeff ); feedforward = K f *neighbor effect ; Valve opening degree = valve opening +feedforward; Among them: neighbor effect The parameter representing the influence of temperature on adjacent rooms; T neighbor Temperature of the adjacent room; T set-neighbor Set the temperature for adjacent rooms; coupling coeff The distance correlation coefficient between adjacent rooms; feedforward is the feedforward control parameter; K f This represents the feedforward control coefficient.

[0034] This invention, when calculating the input to the PID controller, takes into account the different needs of various users. This difference can be reflected by adjusting the target temperature or directly modifying the proportional gain of the PID controller. For example, for users with larger areas or in unfavorable locations (such as top floors or corner units), their target temperature can be appropriately increased or the proportional gain can be amplified to provide more heat compensation. This invention not only considers the impact of area size on heat demand but also pays special attention to the impact of location factors on thermal efficiency, achieving more personalized heating management. Based on real-time feedback, the target flow rate is dynamically adjusted: unlike traditional fixed flow rate distribution methods, this system can dynamically adjust the target flow rate for each household based on real-time monitoring data, ensuring optimal temperature balance under different environmental conditions.

[0035] Example 2: This embodiment optimizes S3 based on Embodiment 1 and proposes a centralized heating heat balance method, including the following steps: S1 uses a PID controller to periodically adjust the temperature ΔT based on the static parameters and target values ​​of each household. i Construct the compensation weight W for each household i The static parameters include the area parameter S. i / max(S) and position parameters, the position parameters including the edge parameter is corner And the top floor parameter is top ,but: W i =α*S i / max(S)+β*is corner +γ*is top +λ*PID(ΔT i ); Wherein: S i / max(S) represents the area normalized value of the actual heating area of ​​each household and the maximum heating area among all users. The location parameter takes values ​​of 0 or 1. α is the area coefficient, β is the edge coefficient, γ is the top floor coefficient, λ is the temperature sensitivity coefficient, and ΔT is the area coefficient. i Set temperature T for each household set and the current temperature T actualThe difference. In this invention, the temperature sensitivity coefficient λ needs to be determined based on the system's thermal inertia experiment. The typical value range of λ is between 0.2 and 0.5. In this embodiment, λ = 0.3 is selected; α, β, and γ are all set empirical values. In this embodiment, α = 0.6, β = 0.3, and γ = 0.2.

[0036] The control cycle of the PDI controller is 15-30 minutes, taking thermal inertia into account; the proportional coefficient Kp, integral coefficient Ki, and derivative coefficient Kd of the PDI controller are obtained using the Ziegler-Nichols empirical method, and the PID (ΔT) i This refers to the situation where, under the set proportional coefficient Kp, integral coefficient Ki, and differential coefficient K, the temperature difference is ΔT. i At that time, the output value of the PDI controller.

[0037] PID Controller Design: Each user's balancing valve heat meter can be considered an independent controlled object. Its input is the target flow rate command issued by the host computer, and its output is the actual flow rate achieved. This invention uses a PID control strategy to dynamically adjust the flow rate for each user, so that the actual temperature of each user is as close as possible to the target temperature.

[0038] The PID controller parameters are set as follows: Proportional term (P): Adjusted directly based on the current error (i.e., the difference between the target temperature and the actual temperature) to quickly respond to temperature changes.

[0039] Integral term (I): Considers the cumulative error over a long period of time, helps to eliminate static error, and is especially suitable for families that are difficult to control precisely through proportional action. It sets the dead zone range to not accumulate integrals and performs anti-saturation treatment of the integral term.

[0040] Differential term (D): Predicts future error trends and prevents overshoot and oscillations, which is especially important when temperatures change rapidly.

[0041] The specific steps of the Ziegler-Nichols empirical method of this invention are as follows: (1) First set Ki = 0, Kd = 0, and gradually increase Kp until the system exhibits small oscillations; (2) Record Ku and the oscillation period Tu at this time; (3) Calculate the initial values ​​of the PID parameters according to the table above; (4) Fine-tune the parameters until the temperature fluctuation is <±0.3℃; (5) Output Kp, Ki and Kd at this time. In this embodiment, Kp = 0.1, Ki = 0.01, Kd = 0.05; The following is an example of dynamic adjustment calculation for PID controller parameters: For the parameters of a top-floor user in a certain unit S i =90m 2 max(S) = 120m 2 is corner =1, is top =1 T set =22℃, T actual =20℃ to obtain ΔT=T set -T actual = +2℃ Static weight calculation: static_weight=α*Si / max(S)+β*iscorner+γ*istop=0.6*(90 / 120)+0.3*1+0.2*1=0.45+0.3+0.2=0.95 PID controller calculation (Kp = 0.1, Ki = 0.01, Kd = 0.05) pid=PIDController(0.1,0.01,0.05) temp_correction=λ*PID(ΔTi)=0.3*pid(2)=0.25; W i =α*Si / max(S)+β*iscorner+γ*istop+λ*PID(ΔTi)=0.95+0.25=1.2; Personalized PID parameter settings: The proportional, integral, and derivative gains of the PID controller can be customized to suit the specific circumstances of different households, such as room size and location characteristics. This customized approach can more effectively handle various complex situations, improving system response speed and stability.

[0042] Integrated PID control loop design: Each household is equipped with an independent PID controller, which monitors the indoor temperature in real time and dynamically adjusts the opening of the balancing valve to achieve the target flow rate, thereby achieving precise temperature control.

[0043] S2, obtain the weight normalization parameter matrix all for all users. weights = [W1, W2, ..., W i ]; S3, according to all weights Calculate the target traffic Qtarget for each household. i S3 specifically includes the following steps: S31, according to all weights Calculate the total weight and sum weights =sum(all) weights), for all weights Summation; in this embodiment, all is obtained by substituting the above parameters. weights = [1.2, 0.8, 1.1...]; S32, calculate the target flow for each household: Qtarget i =Q total *(W i / sum weights ), where: Q total =1000m 3 / h represents the total flow rate of the centralized heating system.

[0044] Preferably, when ΔT i When the value is greater than 0, heating is required. A smoothing factor K is introduced to prevent oscillations, and the flow distribution is increased proportionally. At this time: Qtarget i =Q total *(W i / sum weights )*(1+K*ΔT i ).

[0045] S33, determine if the user is an end user; if so, calculate the end user's Qtarget. i Increase the base traffic margin by 5%-10%, meaning the target traffic range for end users is (Qtarget). i *105%, Qtarget i *110%), the purpose is to provide special compensation to marginal users, and then output the target traffic for each household; otherwise, directly output the target traffic for each household.

[0046] S34, determine the total target traffic ΣQtarget for all users. i Is it greater than Q? total If so, the target traffic for each household is equal to Qtarget. i *(Q total / ΣQtarget i When the total demand exceeds the network capacity, the target flow for each household is compressed proportionally; otherwise, the target flow for each household is output.

[0047] S4, according to Qtarget i Controlling the opening degree of each household's valve; S4 specifically includes the following steps: S41, Valve opening change: valve opening =(Qtarget) i / Qmax)*100%, where: Qmax is the valve's maximum flow rate; S42 uses a PID controller to adjust the valve based on the valve opening degree. S43, determine whether the current valve has reached the limit position. If so, the PID controller stops accumulating the integral term, dynamically adjusts the integral limit value to prevent the integral term of the PID controller from being too large, and records the over-limit time. S43, determine if the over-limit time is greater than 30 minutes. If yes, issue an over-limit alarm. If no, return to S42.

[0048] Preferably, considering the influence of temperature changes in adjacent rooms, feedforward control is introduced to compensate for temperature changes in adjacent rooms, as follows: neighbor effect =sum((T) neighbor -T set-neighbor *coupling coeff ); feedforward = K f *neighbor effect ; Valve opening degree = valve opening +feedforward; Among them: neighbor effect The parameter representing the influence of temperature on adjacent rooms; T neighbor Temperature of the adjacent room; T set-neighbor Set the temperature for adjacent rooms; coupling coeff The distance correlation coefficient between adjacent rooms; feedforward is the feedforward control parameter; K f This represents the feedforward control coefficient.

[0049] This invention, when calculating the input to the PID controller, takes into account the different needs of various users. This difference can be reflected by adjusting the target temperature or directly modifying the proportional gain of the PID controller. For example, for users with larger areas or in unfavorable locations (such as top floors or corner units), their target temperature can be appropriately increased or the proportional gain can be amplified to provide more heat compensation. This invention not only considers the impact of area size on heat demand but also pays special attention to the impact of location factors on thermal efficiency, achieving more personalized heating management. Based on real-time feedback, the target flow rate is dynamically adjusted: unlike traditional fixed flow rate distribution methods, this system can dynamically adjust the target flow rate for each household based on real-time monitoring data, ensuring optimal temperature balance under different environmental conditions.

[0050] Example 3: This embodiment optimizes S1 based on Embodiment 2 and proposes a centralized heating heat balance method, including the following steps: S1 uses a PID controller to periodically adjust the temperature ΔT based on the static parameters and target values ​​of each household. i Construct the compensation weight W for each household i The static parameters include the area parameter S. i / max(S) and position parameters, the position parameters including the edge parameter is corner And the top floor parameter is top ,but: W i =α*S i / max(S)+β*is corner +γ*is top +λ*PID(ΔT i ); Wherein: S i / max(S) represents the area normalized value of the actual heating area of ​​each household and the maximum heating area among all users. The location parameter takes values ​​of 0 or 1. α is the area coefficient, β is the edge coefficient, γ is the top floor coefficient, λ is the temperature sensitivity coefficient, and ΔT is the area coefficient. i Set temperature T for each household set and the current temperature T actual The difference. In this invention, the temperature sensitivity coefficient λ needs to be determined based on the system's thermal inertia experiment. The typical value range of λ is between 0.2 and 0.5. In this embodiment, λ = 0.3 is selected; α, β, and γ are all set empirical values. In this embodiment, α = 0.6, β = 0.3, and γ = 0.2.

[0051] The control cycle of the PDI controller is 15-30 minutes, taking thermal inertia into account; the proportional coefficient Kp, integral coefficient Ki, and derivative coefficient Kd of the PDI controller are obtained using the Ziegler-Nichols empirical method, and the PID (ΔT) i This refers to the situation where, under the set proportional coefficient Kp, integral coefficient Ki, and differential coefficient K, the temperature difference is ΔT. i At that time, the output value of the PDI controller.

[0052] PID Controller Design: Each user's balancing valve heat meter can be considered an independent controlled object. Its input is the target flow rate command issued by the host computer, and its output is the actual flow rate achieved. This invention uses a PID control strategy to dynamically adjust the flow rate for each user, so that the actual temperature of each user is as close as possible to the target temperature.

[0053] The PID controller parameters are set as follows: Proportional term (P): Adjusted directly based on the current error (i.e., the difference between the target temperature and the actual temperature) to quickly respond to temperature changes.

[0054] Integral term (I): Considers the cumulative error over a long period of time, helps to eliminate static error, and is especially suitable for families that are difficult to control precisely through proportional action. It sets the dead zone range to not accumulate integrals and performs anti-saturation treatment of the integral term.

[0055] Differential term (D): Predicts future error trends and prevents overshoot and oscillations, which is especially important when temperatures change rapidly.

[0056] The specific steps of the Ziegler-Nichols empirical method of this invention are as follows: (1) First set Ki = 0, Kd = 0, and gradually increase Kp until the system exhibits small oscillations; (2) Record Ku and the oscillation period Tu at this time; (3) Calculate the initial values ​​of the PID parameters according to the table above; (4) Fine-tune the parameters until the temperature fluctuation is <±0.3℃; (5) Output Kp, Ki and Kd at this time. In this embodiment, Kp = 0.1, Ki = 0.01, Kd = 0.05; The following is an example of dynamic adjustment calculation for PID controller parameters: For the parameters of a top-floor user in a certain unit S i =90m 2 max(S) = 120m 2 is corner =1, is top =1 T set =22℃, T actual =20℃ to obtain ΔT=T set -T actual = +2℃ Static weight calculation: static_weight=α*Si / max(S)+β*iscorner+γ*istop=0.6*(90 / 120)+0.3*1+0.2*1=0.45+0.3+0.2=0.95 PID controller calculation (Kp = 0.1, Ki = 0.01, Kd = 0.05) pid=PIDController(0.1,0.01,0.05) temp_correction=λ*PID(ΔTi)=0.3*pid(2)=0.25; W i=α*Si / max(S)+β*iscorner+γ*istop+λ*PID(ΔTi)=0.95+0.25=1.2; Personalized PID parameter settings: The proportional, integral, and derivative gains of the PID controller can be customized to suit the specific circumstances of different households, such as room size and location characteristics. This customized approach can more effectively handle various complex situations, improving system response speed and stability.

[0057] Integrated PID control loop design: Each household is equipped with an independent PID controller, which monitors the indoor temperature in real time and dynamically adjusts the opening of the balancing valve to achieve the target flow rate, thereby achieving precise temperature control.

[0058] Considering inertia delay compensation, a Smith predictor is connected to the PID controller. The Smith predictor uses the predicted values ​​to calculate the temperature deviation. ΔT i =T set -predicted T ; predicted T =T actual +(Q current *heat transfer _ coeff )*delay time ; Among them: predicted T This is the predicted temperature value; Q current Current flow rate; heat transfer _ coeff The heating coefficient; delay time This is a delay time.

[0059] Taking into account the diurnal temperature range, the set temperature is dynamically adjusted. It is then determined whether the current time is the nighttime set value; if so, the set temperature is adjusted to T. set =T set-daytime -T0: Lower the set temperature at night; otherwise, maintain the set temperature at T. set =T set-daytime ; where: T set-daytime Set the temperature value for daytime.

[0060] S2, obtain the weight normalization parameter matrix all for all users. weights = [W1, W2, ..., W i ]; S3, according to all weights Calculate the target traffic Qtarget for each household. i S3 specifically includes the following steps: S31, according to all weights Calculate the total weight and sum weights =sum(all) weights ), for all weights Summation; in this embodiment, all is obtained by substituting the above parameters. weights = [1.2, 0.8, 1.1...]; S32, calculate the target flow for each household: Qtarget i =Q total *(W i / sum weights ), where: Q total =1000m 3 / h represents the total flow rate of the centralized heating system.

[0061] Preferably, when ΔT i When the value is greater than 0, heating is required. A smoothing factor K is introduced to prevent oscillations, and the flow distribution is increased proportionally. At this time: Qtarget i =Q total *(W i / sum weights )*(1+K*ΔT i ).

[0062] S33, determine if the user is an end user; if so, calculate the end user's Qtarget. i Increase the base traffic margin by 5%-10%, meaning the target traffic range for end users is (Qtarget). i *105%, Qtarget i *110%), the purpose is to provide special compensation to marginal users, and then output the target traffic for each household; otherwise, directly output the target traffic for each household.

[0063] S34, determine the total target traffic ΣQtarget for all users. i Is it greater than Q? total If so, the target traffic for each household is equal to Qtarget. i *(Q total / ΣQtarget i When the total demand exceeds the network capacity, the target flow for each household is compressed proportionally; otherwise, the target flow for each household is output.

[0064] S4, according to Qtarget i Controlling the opening degree of each household's valve; S4 specifically includes the following steps: S41, Valve opening change: valveopening =(Qtarget) i / Qmax)*100%, where: Qmax is the valve's maximum flow rate; S42 uses a PID controller to adjust the valve based on the valve opening degree. S43, determine whether the current valve has reached the limit position. If so, the PID controller stops accumulating the integral term, dynamically adjusts the integral limit value to prevent the integral term of the PID controller from being too large, and records the over-limit time. S43, determine if the over-limit time is greater than 30 minutes. If yes, issue an over-limit alarm. If no, return to S42.

[0065] Preferably, considering the influence of temperature changes in adjacent rooms, feedforward control is introduced to compensate for temperature changes in adjacent rooms, as follows: neighbor effect =sum((T) neighbor -T set-neighbor *coupling coeff ); feedforward = K f *neighbor effect ; Valve opening degree = valve opening +feedforward; Among them: neighbor effect The parameter representing the influence of temperature on adjacent rooms; T neighbor Temperature of the adjacent room; T set-neighbor Set the temperature for adjacent rooms; coupling coeff The distance correlation coefficient between adjacent rooms; feedforward is the feedforward control parameter; K f This represents the feedforward control coefficient.

[0066] This invention, when calculating the input to the PID controller, takes into account the different needs of various users. This difference can be reflected by adjusting the target temperature or directly modifying the proportional gain of the PID controller. For example, for users with larger areas or in unfavorable locations (such as top floors or corner units), their target temperature can be appropriately increased or the proportional gain can be amplified to provide more heat compensation. This invention not only considers the impact of area size on heat demand but also pays special attention to the impact of location factors on thermal efficiency, achieving more personalized heating management. Based on real-time feedback, the target flow rate is dynamically adjusted: unlike traditional fixed flow rate distribution methods, this system can dynamically adjust the target flow rate for each household based on real-time monitoring data, ensuring optimal temperature balance under different environmental conditions.

[0067] Example 4: Based on Examples 1-3, the empirical values ​​α, β, and γ in this embodiment can also be updated periodically using a big data platform. The periodic update process of the big data platform is as follows: S11, Data Collection and Preprocessing: This requires collecting all relevant data, including but not limited to: Area per household: The actual heating area of ​​the household; Location information of the property: factors such as the orientation of the resident's residence, such as whether it is a corner unit or a top floor unit; Current indoor temperature: The typical indoor temperature is collected in real time by multiple temperature sensors placed indoors; Target temperature: Users can set a suitable indoor temperature and report the set temperature value; Current flow rate or opening degree of the balancing valve heat meter: Real-time acquisition of relevant parameters of the balancing valve heat meter; Meteorological temperature collection: Meteorological data are collected regularly.

[0068] S12: The big data platform acquires all relevant data, uses real-time heating network data to assess system energy efficiency, loads the building thermodynamics model, updates the global strategy, optimizes offline parameters, and then uses online reinforcement learning to obtain empirical values ​​α, β, and γ.

[0069] Intelligent initialization phase: S121 utilizes a big data platform to comprehensively analyze data from various sources, including historical heating records, building structure data, user lifestyle preferences, and weather forecast information.

[0070] S122 uses AI algorithms to conduct detailed needs analysis of families in the region.

[0071] S123 dynamically generates more accurate personalized initial demand weights based on AI analysis results.

[0072] S124 sends data-driven, personalized initial flow rate commands to the smart balance valve heat meter in each household to achieve a high starting point for balancing.

[0073] AI-enhanced real-time operation phase: SA1 uses an IoT sensor network to collect real-time indoor temperature and environmental parameters of each household, forming a big data stream for heating.

[0074] The SA2 core controller not only uses traditional PID control, but also employs AI optimization algorithms to perform millisecond-level analysis of massive real-time data.

[0075] SA3's AI algorithm comprehensively considers factors such as individual household needs, neighboring household influences, pipeline status, and heat source efficiency to calculate a more accurate and coordinated AI-optimized flow rate adjustment.

[0076] SA4 instantly sends AI-generated optimized flow rate commands to the corresponding smart balance valve heat meter, enabling dynamic, efficient, and precise heating control.

[0077] Big data-driven intelligent feedback and evolution mechanism: SB1, the big data platform, continuously gathers all operational data from the system, including temperature, flow rate, energy consumption, comfort feedback (if accessible), and external environment for each household.

[0078] SB2, the AI ​​analytics engine (machine learning / deep learning model), periodically (or in near real-time) evaluates key performance indicators of the overall system (such as average temperature comfort deviation, maximum temperature deviation, energy efficiency, system stability, etc.).

[0079] Based on big data insights and AI diagnostics, the SB3 system is able to: Adaptive adjustment: Automatically fine-tunes the weight of each household's individual needs to achieve precise load allocation.

[0080] Intelligent parameter optimization: Automatically updates and optimizes the PID parameters and even model structure of each loop (such as dynamically switching control strategies) to improve single-point and group control performance.

[0081] Continuous Iteration and Evolution: Using historical big data to train and iterate AI models to continuously improve prediction accuracy, control strategies, and system balance.

[0082] Global collaborative optimization: Identify bottlenecks and optimization potential in the heating network to guide the adjustment of output of heat source stations or the overall network collaborative optimization strategy.

[0083] Regular performance evaluation and parameter readjustment: The system has a self-evaluation function, which can regularly evaluate the overall performance based on actual operating data, and readjust the weight factors or PID parameters of each household accordingly to continuously optimize the system's balance and efficiency.

[0084] Intelligent feedback mechanism: An effective feedback mechanism has been established to collect and analyze the actual operating data of each household, promptly identify and resolve potential problems, and ensure the long-term stable and efficient operation of the system.

[0085] Intuitive user interface: It provides a user-friendly management interface, which allows administrators to easily view the real-time status, historical data and system performance indicators of each household, facilitating necessary manual intervention or adjustments.

[0086] Flexible emergency response plans: Flexible emergency plans have been developed for special situations (such as extreme weather conditions), which enhances the system's adaptability and reliability.

Claims

1. A method for achieving heat balance in centralized heating, characterized in that, Includes the following steps: S1 uses a PID controller to periodically adjust the temperature ΔT based on the static parameters and target values ​​of each household. i Construct the compensation weight W for each household i ; S2, obtain the weight normalization parameter matrix all for all users. weights = [W1, W2, ..., W i ]; S3, according to all weights Calculate the target traffic Qtarget for each household. i ; S4, according to Qtarget i Control the opening degree of the valve for each household; The static parameters include the area parameter S. i / max(S) and position parameters, the position parameters including the edge parameter is corner And the top floor parameter is top ,but: W i =a*S i / max(S)+β*is corner +γ*is top + λ*PID(ΔT i ); Wherein: S i / max(S) represents the area normalized value of the actual heating area of ​​each household and the maximum heating area among all users. The location parameter takes values ​​of 0 or 1. α is the area coefficient, β is the edge coefficient, γ is the top floor coefficient, λ is the temperature sensitivity coefficient, and ΔT i Set temperature T for each household set and the current temperature T actual The difference.

2. The centralized heating heat balance method according to claim 1, characterized in that, S3 specifically includes the following steps: S31, according to all weights Calculate the total weight and sum weights =sum(all weights ), for all weights Summation; S32, calculate the target flow for each household: Qtarget i =Q total *(W i / sum weights ), where: Q total This represents the total flow rate of the centralized heating system.

3. The centralized heating heat balance method according to claim 1, characterized in that, S4 specifically includes the following steps: S41, Valve opening change: valve opening = (Qtarget i / Qmax) * 100%, where: Qmax is the valve's maximum flow rate; S42 uses a PID controller to adjust the valve based on the valve opening degree. S43, determine whether the current valve has reached its limit position. If so, the PID controller stops accumulating the integral term and records the over-limit time. S43, determine if the over-limit time is greater than 30 minutes. If yes, issue an over-limit alarm. If no, return to S42.

4. The centralized heating heat balance method according to claim 2, characterized in that, S3 further includes the following steps: S33, determine if the user is an end user; if so, calculate the end user's Qtarget. i Increase the base traffic margin by 5%-10% before outputting the target traffic for each household; otherwise, directly output the target traffic for each household.

5. A centralized heating heat balance method according to any one of claims 1-4, characterized in that, S3 further includes the following steps: S34, determine the total target traffic ΣQtarget for all users. i Is it greater than Q? total If so, the target traffic for each household is equal to Qtarget. i *(Q) total / ΣQtarget i If not, output the target traffic for each household.

6. The method for achieving heat balance in centralized heating according to claim 1, characterized in that, The control cycle of the PDI controller is 15-30 minutes; the proportional coefficient Kp, integral coefficient Ki, and derivative coefficient Kd of the PDI controller are obtained using the Ziegler-Nichols empirical method, and the PID (ΔT) i This refers to the situation where, under the set proportional coefficient Kp, integral coefficient Ki, and differential coefficient K, the temperature difference is ΔT. i At that time, the output value of the PDI controller.

7. A centralized heating heat balance method according to claim 2, characterized in that, When ΔT i When the value is greater than 0, a smoothing factor K is introduced to proportionally increase the flow allocation. At this time: Qtarget i =Q total *(W i / sum weights )* (1+K*ΔT i ) 。 8. A method for achieving heat balance in centralized heating according to claim 1, characterized in that, S1 further includes the following steps: Considering inertia delay compensation, a Smith predictor is connected to the PID controller. The Smith predictor uses the predicted values ​​to calculate the temperature deviation. ΔT i =T set - predicted T ; predicted T =T actual + (Q current * heat transfer _ coeff ) * delay time ; Among them: predicted T This is the predicted temperature value; Q current Current flow rate; heat transfer _ coeff The heating coefficient; delay time This is the delay time.

9. A centralized heating heat balance method according to claim 3, characterized in that, S4 further includes the following steps: Considering the impact of temperature changes in adjacent rooms, feedforward control is introduced to compensate for temperature changes between adjacent rooms, as detailed below: neighbor effect = sum( (T neighbor - T set-neighbor ) * coupling coeff ); feedforward = K f * neighbor effect ; Valve opening = valve opening +feedforward; Among them: neighbor effect The parameter representing the influence of temperature on adjacent rooms; T neighbor Temperature of the adjacent room; T set-neighbor Set the temperature for adjacent rooms; coupling coeff The distance correlation coefficient between adjacent rooms; feedforward is the feedforward control parameter; K f This represents the feedforward control coefficient.

10. A method for achieving heat balance in centralized heating according to claim 1, characterized in that, S1 further includes the following steps: Taking into account the diurnal temperature range, the set temperature is dynamically adjusted. It is then determined whether the current time is the nighttime set value; if so, the set temperature is adjusted to T. set = T set-daytime -T0, if not, set the temperature to remain at T. set = T set-daytime ; where: T set-daytime Set the temperature value for daytime.

Images