阅读说明：本技术 基于互联网的加热控制参数优化的方法和装置 (Heating control parameter optimization method and device based on Internet ) 是由 扬·施特鲁贝尔 克里斯蒂安·阿尔诺德 于 2018-02-05 设计创作，主要内容包括：本发明涉及一种用于确定暖通空调和制冷(HVACR)系统(2)的闭环控制器(3)或开环控制器的最优的控制参数集(Θ<Sub>k</Sub>)的方法。在第一方法步骤中,室外温度(T<Sub>A</Sub>),房间(9)的实际室内温度(T<Sub>R</Sub>),供应温度(T<Sub>VL</Sub>),预先确定的目标室内温度(T<Sub>R,W</Sub>),和预先确定的目标供应温度(T<Sub>VL,W</Sub>)均被检测。从检测到的测量值(T<Sub>A</Sub>,T<Sub>R</Sub>,T<Sub>R,W</Sub>,T<Sub>VL</Sub>,T<Sub>VL,W</Sub>)和检测数据包(D<Sub>k</Sub>)的时间(t<Sub>k</Sub>)的生成,并通过互联网连接传输到服务器(8),其中数据包(D<Sub>k</Sub>)存储在连接到服务器(8)的存储介质(6、7)中。在下一个方法步骤中,最优的控制参数集(Θ<Sub>k</Sub>)基于传输和存储的所述数据包(D<Sub>k</Sub>)的测量值(T<Sub>A</Sub>,T<Sub>R</Sub>,T<Sub>R,W</Sub>,T<Sub>VL</Sub>,T<Sub>VL,W</Sub>),基于在特定时间段(Δt)的较早时间(t...k)生成的多个其他数据包(D<Sub>0...k</Sub>)的测量值(T<Sub>A</Sub>,T<Sub>R</Sub>,T<Sub>R,W</Sub>,T<Sub>VL</Sub>,T<Sub>VL,W</Sub>)；和/或通过在服务器(8)上执行计算算法的多个先前确定的最优的控制参数集合(Θ<Sub>k-1</Sub>)中的至少一个来计算。在接下来的方法步骤中,计算的最优的控制参数集(Θ<Sub>k</Sub>)存储在连接到所述服务器(8)的所述存储介质(6,7)中,并通过互联网连接传输到计算的最优的控制参数集(Θ<Sub>k</Sub>)存储在连接到所述服务器(8)的所述存储介质(6,7)中,或传输到HVACR系统(2)的用户(B)。(The invention relates to a method for determining an optimal control parameter set (Θ) for a closed-loop controller (3) or an open-loop controller of a heating, ventilation, air conditioning and refrigeration (HVACR) system (2) k ) The method of (1). In a first method step, the outdoor temperature (T) A ) Actual room temperature (T) of the room (9) R ) Supply temperature (T) VL ) A predetermined target indoor temperature (T) R,W ) And a predetermined target supply temperature (T) VL,W ) Are all detected. From the detected measured value (T) A ,T R ,T R,W ,T VL ,T VL,W ) And detecting the data packet (D) k ) Time (t) of k ) And is transmitted through an internet connectionTo a server (8) in which data packets (D) are transmitted k ) Are stored in storage media (6,7) connected to a server (8). In the next method step, the optimal control parameter set (Θ) k ) Based on said data packet (D) transmitted and stored k ) Measured value (T) of A ,T R ,T R,W ,T VL ,T VL,W ) Based on a plurality of further data packets (D) generated at an earlier time (t.. k) of a specific time period (Δ t) 0...k ) Measured value (T) of A ,T R ,T R,W ,T VL ,T VL,W ) (ii) a And/or by executing a plurality of previously determined optimal sets of control parameters (Θ) of a calculation algorithm on the server (8) k‑1 ) Is calculated. In the next method step, the calculated optimal control parameter set (Θ) k ) Is stored in said storage medium (6,7) connected to said server (8) and transmitted to the calculated optimal control parameter set (Θ) via an internet connection k ) Stored in said storage medium (6,7) connected to said server (8) or transmitted to a user (B) of an HVACR system (2).)

1. A control parameter set (Θ) for determining an optimal control parameter set (Θ) for a closed-loop controller (3) or an open-loop controller of a heating, ventilation, air conditioning and refrigeration (HVACR) system (2)_k) The method comprises the following steps:

detecting outdoor temperature (T)_A)；

Detecting the actual indoor temperature (T) of the room (9)_R)；

Detecting the supply temperature (T)_VL)；

Detecting a predetermined target indoor temperature (T)_R,W)；

Detecting a predetermined target supply temperature (T)_VL,W)；

Generating a measured value (T) with said detection_A,T_R,T_R,W,T_VL,T_VL,W) And a detection time (t)_k) Data packet (D)_k)；

Transmitting the data packet (D)_k) To a server (8) via an internet connection;

transmitting the data packet (D)_k) Stored in a storage medium (6,7) connected to the server (8);

according to said data packet (D) transmitted and stored_k) Measured value (T) of_A,T_R,T_R,W,T_VL,T_VL,W) And calculating an optimal control parameter set based on the following parameters (Θ_k)：

A plurality of further data packets (D.) generated at an earlier time (t.. k) of a specific time period (Δ t)_0...k) Measured value (T) of_A,T_R,T_R,W,T_VL,T_VL,W) (ii) a And/or

A plurality of previously determined optimal control parameter sets (Θ)_k-1) At least one of;

wherein the optimal control parameter set (Θ) is calculated by executing a calculation algorithm on the server (8)_k)；

The optimal control parameter set (Θ) to be calculated_k) Is stored in said storage medium (6,7) connected to said server (8); and

setting the calculated optimal control parameter set (Θ)_k) Transmitted from the server (8) to the closed-loop controller (3) or the open-loop controller of the HVACR system (2) over the Internet connection, or with the calculated optimal control parameter set (Θ)_k) Sending a notification to a user (B) of the HVACR system (2).

2. The method of claim 1, wherein the method further comprises:

at said time (t)_k) Detecting at least one further value of the following composition:

reflux temperature (T)_RL)；

Mass flow rate

Solar radiation (G)_sol)；

External heat source input (P)_FW)；

Wherein the generated data packet (D)_k) Values comprising said further measurements

3. Method according to claim 1 or 2, wherein the method is used for calculating the optimal control parameter set (Θ)_k) Is based on the data packets (D) stored in the storage medium (6,7)_0...k) Is predetermined.

4. The method according to at least one of the preceding claims, wherein the optimal control parameter set (Θ)_k) The calculating of (a) includes:

for each data packet (D) of a predetermined time period (Delta t)_0...k) Determination of the actual Room temperature (T)_R) With target indoor temperature (T)_R,W) First deviation (e) therebetween_R)；

For each data packet (D) of a predetermined time period (Delta t)_0...k) Determination of supply Room temperature (T)_R) With a predetermined supply temperature (T)_VL,W) Second deviation (e) therebetween_VL) (ii) a And

deviation (e) based on said determination_R,e_VL) Determining and applying a weighting factor between 0 and 1 such that when calculating the optimal control parameter set (Θ)_k) In the case of a high-deviation packet (D), a low-deviation packet (D0... k) is considered more frequently than a high-deviation packet (D)_0...k) Are less considered; or

From stored data packets (D)_0...k) Calculating the optimal control parameter set (Θ)_k) Such that said determined first deviation (e)_R) And/or said determined second deviation (e)_VL) Less than or equal to a respectively predetermined threshold value (delta)_R,δ_VL)。

5. Method according to at least one of the preceding claims, wherein the measured values are detected periodically after a predetermined time interval has elapsed and are repeatedly detected each time a predetermined number of predetermined time intervals have elapsedCalculating the optimal control parameter set (Θ)_k) So that it is used to calculate the optimal control parameter set (Θ)_k) Of a plurality of corresponding generated data packets (D)_0...k) Are stored in said storage medium (6,7) of said server (8).

6. Optimized control parameter set (Θ) for a closed-loop controller (3) or an open-loop controller of a heating, ventilation and cooling (HVACR) system (2)_k) The system (1) of (a), comprising:

for detecting external temperature (T)_A) An external temperature sensor (11);

for detecting the actual indoor temperature (T)_R) Wherein the indoor temperature sensor (10) is disposed within a room (9);

for detecting supply temperature (T)_A) A supply temperature sensor (12);

for presetting a target indoor temperature (T)_R,W) The device (17);

for presetting a target supply temperature (T)_VL,W) The device (4);

for producing measured values (T) with said detection_A,T_R,T_R,W,T_VL,T_VL,W) And a detection time (t)_k) Data packet (D)_k) The device (22) of (a);

transmission means having means for transmitting said data packets (D) via an internet connection_k) An interface (20) for transmission to a server (8); and

a storage medium (6,7) connected to the server (8) for storing the data package (D)_k)；

Wherein the server (8) has a processor (CPU) configured to:

based on the time point (t)_k) Data packet (D) for transmission and storage_k) Measured value (T) of_A,T_R,T_R,W,T_VL,T_VL,W) Calculating an optimal control parameter set (Θ k) and based on:

at an earlier time (t) of a certain time period (Δ t)_0...k-1) A plurality of further ones of the generationData packet (D)_0...k-1) Measured value (T) of_A,T_R,T_R,W,T_VL,T_VL,W) (ii) a And/or

A plurality of previously determined control parameter sets (Θ)_k-1) At least one of (a);

wherein the processor (CPU) is configured to execute a program for calculating the optimal control parameter set (Θ)_k) The calculation algorithm of (1);

setting the calculated optimal control parameter set (Θ)_k) Stored in said storage medium (6,7) connected to a server (8); and

setting the calculated optimal control parameter set (Θ)_k) Transmitted to the closed-loop controller (3) or the open-loop controller of the HVACR system (2) over the Internet over an interface (21), or with the calculated optimal control parameter set (Θ)_k) Sending a notification to a user (B) of the HVACR system (2).

7. The system (1) according to claim 6, wherein the system (1) further comprises:

for determining the reflux temperature (T)_RL) A return temperature sensor (15); and/or

For measuring mass flow

For detecting solar radiation (G)_sol) A solar radiation sensor (14); and/or

For detecting external heat source input (P)_FW) The apparatus of (1);

wherein the means (22) for generating a data packet is configured to generate a data packet having the further measurement value

8. System (1) according to claim 6 or 7, wherein the processor (CPU) of the server (8) is configured for storing the data packet (D) based on_0...k) Is predetermined for calculating said optimal control parameter set (Θ k).

9. The system (1) according to any one of claims 6 to 8, wherein the processor (CPU) of the server (8) is configured for:

for each data packet (D) of a predetermined time period (Delta t)_0...k) Determination of the actual Room temperature (T)_R) With target indoor temperature (T)_R,W) A first deviation (eR) therebetween;

for each data packet (D) of a predetermined time period (Delta t)_0...k) Determination of supply Room temperature (T)_R) With a predetermined supply temperature (T)_VL,W) Second deviation (e) therebetween_VL) (ii) a And

deviation (e) based on said determination_R,e_VL) Determining and applying a weighting factor between 0 and 1 such that when calculating the optimal control parameter set (Θ)_k) Time, low deviation data packet (D)_0...k) High-deviation data packets (D) are considered more_0...k) Are less considered; or

From stored data packets (D)_0...k) Determining the optimal control parameter set (Θ)_k) Such that said determined first deviation (e)_R) And/or said determined second deviation (e)_VL) Less than or equal to respective predetermined threshold values (δ R, δ VL).

10. The system (1) according to any one of claims 6 to 9, wherein the sensor detects the measurement value periodically after a predetermined time interval has elapsed, and the processor (CPU) of the server (8) is configured for repeating for a predetermined plurality of predetermined times each timeCalculating the optimal control parameter set (Θ)_k) So that a corresponding plurality of generated data packets (D)_0...k) Is stored in the storage medium (6,7) of the server (8) for calculating the optimal control parameter set (Θ)_k)。

Technical Field

The invention relates to a method for optimizing control parameters of a closed-loop controller or an open-loop controller of an Internet-based heating, ventilation, air conditioning and refrigeration (HVACR) system, in particular to a central heating system for a building. The invention also relates to a system for performing the method according to the invention.

Background

Typically, the closed-loop controller or open-loop controller of an HVACR system operates with open-loop and/or closed-loop control algorithms parameterized by the user, such as a heating contractor or end customer during commissioning. Non-optimal parameterization, i.e. non-optimal control parameters using open-loop and/or closed-loop control algorithms, may result in excessive energy consumption. On the other hand, at lower outside temperatures, for example, too little heat may be provided to the building. Therefore, it is conceivable that firepower may be provided too little to cool the building at a higher outside temperature. In this case, the user of the HVACR system can further adjust (iterate as necessary) the control parameters according to the respective dynamic conditions to adapt to different application conditions, for example depending on the building dynamics, in order to improve the control quality, reduce the operating costs or increase the efficiency of the system. However, in order to optimize the control parameters, the user must have expertise in the process to be controlled and the effect of the algorithm parameters on the operation of the system. As an alternative to manually adjusting the control parameters of the open-loop and/or closed-loop control algorithms, an automatic adjustment may be performed using a superimposed identification and adaptation method. In this way the heating contractor and the end-user are completely free from the task of optimizing the parameter settings. In this way, for example, an adaptive heating characteristic of the heating circuit can be achieved.

According to the prior art, weather-guided control of a heating system, i.e. a control system operating according to weather or season, can ensure that the supply temperature of the heating circuit is adjusted to the heat demand of the building according to the outside temperature. If the heating system is set in an optimal way, only the currently required amount of heat is generated. For this purpose, for example, the measurement of the outside temperature, and depending on the desired room temperature (target room temperature) and the boundary conditions of the building, the supply temperature required to reach the target room temperature is determined. The relationship between the external temperature and the supply temperature is described, for example, by a heating characteristic (also referred to as a heating curve). When the external temperature is low, the heating characteristics result in a higher supply temperature.

An example of the heating characteristics is shown in fig. 3. Thus, the heating characteristics provide a specific target supply temperature for a specific combination of the external temperature and a predetermined target room temperature. The heating characteristics are essentially characterized by two parameters: slope and level. The slope of the heating characteristic indicates the degree to which the supply temperature varies based on the external temperature. In a poorly insulated old house, when the outside becomes cold, the heat loss increases sharply. The heating characteristics must then be very steep to enable the heating to provide sufficient thermal energy at low external temperatures to ensure user comfort. In modern and well insulated houses, changes in the outside temperature have less influence on the heat loss. Thus, the heating characteristics here can be chosen more flat. Therefore, even in very cold weather, a slight increase in supply temperature is sufficient. The level of the heating characteristic defines the base point, i.e., the intersection of the heating characteristic with the vertical axis in fig. 3. By moving the heating characteristics up or down (see heating characteristics c and d), the heat of the heating characteristics can be output. The heating system may be increased or decreased uniformly. For example, if there is always a little too hot in a building, the heating characteristics can be reduced by lowering the level. This lowers the target supply temperature over the entire heating characteristic range. This may reduce the energy consumption of the heating system.

Heating characteristics are often rarely set, usually when the heating system is installed and when the room temperature is repeatedly lowered. This means that the heating system can be operated with higher energy consumption than necessary. The present invention provides a method and corresponding system for optimizing control parameters of a closed-loop controller or an open-loop controller of an HVACR system, i.e. in particular a heating system, over the internet. The method or system according to the invention allows the HVACR system to be adapted to varying boundary conditions caused or affected, for example, by fluctuations in the external temperature, solar radiation or changes in the building itself. This means that HVACR systems can be operated in a particularly efficient and energy efficient manner. In addition, starvation can be avoided.

A method for internet based heating system adaptation is disclosed in e.g. US patent application US2015/0032267 a 1. Sensors connected to the heating control unit can measure the internal and external temperatures. The control unit uses the neural network to determine a prediction of the required heating energy, from which a modified external temperature can be calculated to control the heating control accordingly. Here, the control parameters are not optimized, but the output signal of the controller is changed in some way by modifying the input signal of the controller.

Another control system for building heating is known from german translation DE69918379T2 of the european patent specification. The plurality of sensing elements measure an external temperature as well as a supply temperature and a room temperature. The system includes means for calculating an optimal heat output through a neural network. The optimal heat output is calculated by predicting the external conditions and predicting the internal temperature of the building.

Disclosure of Invention

The method described below according to the invention will optimize the control parameters required for closed-loop or open-loop control of a heating ventilation air conditioning and refrigeration system (HVACR) installed in a building with internet support. The optimization parameters may then be set automatically, or may be set by the user on a closed-loop or open-loop controller. For this purpose, the operating recommendations may be sent to a user of the closed-loop or open-loop controller, for example a heating contractor or an end user of the heating system. The optimized control parameters may be displayed by devices communicating via the internet (e.g., via apps on the mobile terminal device).

To determine and optimize control parameters, a device (e.g., an embedded system) records the relevant system state of the HVACR system using appropriate sensors over a period of time and sends it over an internet connection to a central server for further processing. Relevant system conditions include, for example, the outside temperature of the building, the room temperature of the reference room, the supply temperature, the return temperature, a predetermined target room temperature, a particular target supply temperature, the degree of solar radiation on the building and the external heat source input to the reference room. These system states or system variable portions are detected as measured values by the respective sensors and transmitted as data packets to the server.

The server processes and analyzes the detected system variables to calculate the optimized control parameters. In addition, the server transmits the control parameters optimized by the calculation algorithm to the user. The parameters communicated to the user may be as suggested actions to adjust the optimized control parameters in the controller of the HVACR system. Alternatively, the calculated optimized control parameters may also be directly received and accepted by a closed-loop or open-loop controller of the HVACR system, so that the control of the HVACR system may be automatically optimized. The control parameters may be optimized periodically so that the HVACR system may continually adapt to changing boundary conditions.

Brief description of the drawings

Further advantageous embodiments are described in more detail below on the basis of exemplary embodiments shown in the drawings, to which, however, the invention is not restricted.

These figures show schematically:

fig. 1, fig. 1 shows a structural design of a system or method for optimization of control parameters of an internet based heating system according to a first exemplary embodiment of the present invention;

FIG. 2, FIG. 2a and FIG. 2b show more details of the first exemplary embodiment of the present invention;

FIG. 3, FIG. 3 shows four exemplary heating characteristic curves;

FIG. 4, FIG. 4 shows an example of heating characteristics suitable for multiple measurements;

FIG. 5, FIG. 5a and FIG. 5b show a long time series of room temperature measurements and the corresponding process of optimizing control parameters, respectively;

FIG. 6, FIG. 6a shows a simplified signal flow of the physical process of the heating circuit; fig. 6b shows an approximation of this process by a mathematical model.

Detailed description of the specific embodiments

In the following description of the preferred embodiments of the present invention, the same reference numerals are used for the same or similar components.

First exemplary embodiment

Fig. 1 shows an optimum control parameter set Θ or Θ for calculating an optimum control parameter set Θ or Θ for a closed-loop controller 3 or an open-loop controller 3 of an HVACR (heating, ventilation, air conditioning and refrigeration) system 2 with a characteristic signal flow according to the method of the invention_kAccording to a first exemplary embodiment of (1). Vector quantity

Or

Representing a total of q sets of control parameters. The index k refers to the current point in time and indicates that this is the optimal set of control parameters for the iteration. For example, the control parameter set Θ without the operation index k is used to control the operation index kThere is an iterative computation method in the conventional least squares method.

As an example of HVACR system 2, consider heating system 2 below. The heating system 2 is installed in a building. The following is an example of a single room 9, the temperature T of which_RControlled by the heating system 2. The room 9 can also be used as a reference room for temperature control of the whole building. Alternatively, several reference rooms may be used, and temperature control may be performed for each room in the building.

The heating system 2 continuously detects the relevant system state, e.g. the external temperature T_ARoom temperature T_RSupply temperature T_VLTemperature of reflux T_RLAnd the mass flow of the carrier medium through the supply line of the heating system 2 or through the return lineAnd these measurements are summed to a measurement time t_kData packet D of_kAnd transmits it to the server 8 through the internet. In addition, the solar radiation G contributing to the heating of the building or room can be detected by suitable sensors_sol。

The server 8 is designed to store and evaluate the received data packets D_kAnd thus determine an optimal control parameter set theta for the heating controller 3 of the heating system 2_k. Then, the optimal control parameter set Θ_kMay be sent to user B, e.g. a heating contractor or an end-user, and the set of optimization control parameters Θ_kMay be sent to user B, e.g. a heating contractor or an end customer, which may then decide whether to accept the control parameter Θ_k. In this case, user B sets a control parameter Θ on controller 3_k. Alternatively, the optimal control parameter Θ_kOr may be transmitted directly to and employed by the heating controller 3 via the internet.

The method according to the invention can therefore also be used for heating controllers of older heating systems whose control parameters cannot be adjusted via the internet. Central server 8 may also determine optimal control parameters for several different remote HVACR systems simultaneously. For this, the server receives 8 data packets of a plurality of HVACR systems, evaluates them, and then transmits the calculated optimal control parameters to the corresponding users or controllers.

Fig. 2a and 2b show more details of the first exemplary embodiment of the present invention. FIG. 2a shows the system 1 for determining the optimal control parameter Θ_kPart of the set of (a). The system includes a plurality of sensors for recording the status of the associated system. For detecting the supply temperature T_VLIs arranged in the supply of the heating system 2. In the return flow of the heating system 2, there is a temperature T for detecting the return flow_RLThe return temperature sensor 15. Mass flow sensors may also be disposed in the supply and/or return lines to detect mass flow

An indoor temperature sensor 10 disposed in the room 9 detects an indoor temperature T_R. The heating system 2 provides a carrier medium to the supply flow, which carrier medium transfers heat to a heat sink 16 arranged in the room 9. The thermostat 17 in the room 9 can be used to predetermine the target room temperature T_R，W。

Outside the building, there is a sensor for detecting the outside temperature T_AAnd an external temperature sensor 11 for detecting solar radiation G_solThe solar radiation sensor 14. In addition, at least one sensor for detecting an external heat source input P may be arranged in the room_FWAdditional sensors (not shown). External heat source input P_FWThe heat source of (b) may be, for example, an electric appliance that generates heat, such as a refrigerator or an electric stove.

The controller 3 of the heating system 2 has a heating circuit controller 4 and a boiler controller 5. The heating circuit controller 4 is provided with at least the following measurements: room temperature T_RExternal temperature T_ATarget Room temperature T_R，WAnd a supply temperature T_VL. Optionally, at least one measured variable, i.e. solar radiation G_solMass flow rate

And/or an external heat source input P_FWA heating loop controller may be provided. The heating controller 4 determines the target supply temperature T using a control method predefined for the heating controller 4_VL，WThe target supply temperature T_VL，WIs output to the boiler controller 5. The control method used by the heating controller 4 may be based on temperature, for example, as described in more detail below. Typically, the heating controller 4 uses a set of control parameters

For example, for the heating characteristic, the control parameter set includes, for example, two parameter slopes

And level

FIG. 2b shows a packet D_kTransmission of the measured values in the form to the server 8 for calculating the optimized control parameters theta_k. The heating controller 3 comprises means for generating a data packet D with measured values_kAnd a time point t for detecting the measured value_k. Via the interface 20 to the internet, data packet D_kIs sent to the server 8 which is also connected to the internet via the interface 21.

The server 8 includes: a temporary memory 6 in which the received data packet D is first stored_k(ii) a And a permanent memory 7 having a database in which the received data packets and the calculated control parameters are stored. The server 8 also has a CPU processor configured to calculate an optimal control parameter set Θ from the data packet_k. The optimal set of control parameters Θ can then be passed through the interface 21_kTo the controller 4 or to a user of the heating system 2.

The heating control 4 of the heating system 2 of the first exemplary embodiment can be controlled, for example, by means of a heating characteristic curve. Heating characteristics describe the external temperature T_AAnd a heating system2 supply temperature T_VL，WIs measured. The heating characteristics are defined by two parameter levels theta_NAnd slope theta_SAnd (5) characterizing. Therefore, the heating characteristic parameter set Θ to be optimized_kIs theta_k＝{θ_N，θ_S}_k。

With the heating characteristic curve, the weather guiding operation of the heating system 2 can be achieved. In the optimization method according to the invention, the system state, i.e. the supply temperature T, is detected using the respective temperature sensors 10, 11 and 12 shown in fig. 2a_VLExternal temperature T_AAnd a reference room T_ROf (2) is used. For example, target room temperature T_R，WMay be predetermined by the user via the thermostat 17.

FIG. 3 shows the target room temperature T at 20 deg.C_R，WSlope theta for different values of the parameter_SAnd a horizontal theta_NFour exemplary heating characteristics a to d. Three heating characteristics a to c are only at the parameter slope theta_SWith a difference in the above. It can be seen that when the slope θ_SThe heating characteristics are steeper as one increases (from a to c). Horizontal theta_NA change in the parameter will correspondingly shift the characteristic in the vertical direction. This is illustrated by way of example with two heating characteristic curves c and d. Dependent on the external temperature T_AAnd adjusted room temperature T_R，WThe heating controller 4 calculates the supply temperature T using the heating characteristics_VL，WIs transmitted to the boiler controller 5 as shown in fig. 2 a.

In the case of prior art universal heating controllers, the two parameters of the heating characteristic slope and level can be adjusted, for example, through a configuration menu on the heating controller. Typically, the installer of the heating system sets the parameters according to the properties of the building only when commissioning the heating system. In some cases, the heating controller may even operate at factory set parameters. The parameters are usually adjusted during the heating operation only in case of significant deviations from the intended comfort of the heating system. To avoid heat starvation, the heating system is typically operated at a heating temperature that is higher than necessary. This may result in excessive energy consumption.

The system 1 according to the invention can be used to optimize a parameter of the heating characteristic, the slope theta_SAnd a horizontal theta_N. This means that the heating system 2 can be operated in a particularly efficient manner. Excessive energy demand or insufficient heat supply is avoided.

As described above, the time-varying measured or target values of the relevant state variables are detected at regular intervals and sent to the server 8 as data packets. The measurement values may be represented here as a data vector, which comprises all measurement values in a time series. For example, it may be possible to have a resolution t of 15 minutes over the course of a day_SThe measured values are recorded in order to generate 96 data packets D in 24 hours₁.., D96 and sends it to the server. Data vector

Thus 96 measurements of the external temperature are included:

thus, the server 8 may be at regular intervals t_SThe optimal control parameters are again calculated. In particular, the control parameters may be calculated iteratively. This means that past time point packets or previously calculated control parameters will be used in the calculation of the control parameters.

The advantages of the iterative method will now be illustrated by comparing the conventional method with the iterative calculations shown according to the present invention. The calculation of the optimal set of control parameters can be performed, for example, using known methods. In the following, two exemplary embodiments of the calculation process are described. With the conventional least squares method for optimizing the control parameters of the heating controller, the measurement data must first be recorded for a long time. The upper graph of FIG. 5a shows the room temperature T over a period of 60 days_RExample of a measurement curve of (a). This means that when 96 packets are detected per day, a total of 60 × 96 — 5760 packets will be stored and evaluated when calculating the control parameters. The lower diagram of fig. 5a shows the initial values Θ of the control parameters₀And optimal controlDesired set of system parameters Θ_opt. The conventional method takes 60 days to deliver the results at this point. The large number of data packets also makes the calculation of the control parameters very time consuming.

The advantage of iteratively performing optimal control parameters is described on the basis of fig. 5 b. The upper diagram shows again the room temperature T_RCurves over 60 days. The lower graph correspondingly shows that the theta is based on the same initial value₀And the required optimal control parameter theta_opt.Control parameter curve of (2). In the example shown, the optimal control parameter set is calculated daily, 96 packets each time are sent and stored, starting with the control parameter set calculated the previous day. This means that the amount of data that has to be processed during the calculation is much smaller. In addition, packets can be deleted every day after the optimal control parameters are calculated, so that the memory requirement is much smaller compared to the conventional method. As shown by the progression of the control parameters in the lower graph of fig. 5b, the result is iteratively close to the required optimal set of control parameters. This means that the controller can operate earlier than in the comparative example in fig. 5a using the optimal control parameters. In addition, the computational workload is greatly reduced.

Standard least squares method

A first exemplary embodiment of a calculation algorithm for calculating the optimal set of control parameters uses a least squares method. The traditional least squares method determines the optimal control parameters Θ by solving a system of linear equations:

AΘ＝b (1)

in a manner

The matrices a and b here represent the so-called data matrices of the underlying estimation problem. In general, more data points and thus lines of the data matrix can be used as parameters to be estimated, so that the present system of linear equations is overdetermined. For example, in Stoer/Bursch: this conventional min-flat method is described by nummerische mathematik, press publishing, berlin, 2007, page 250 and below.

Equation (1) first discloses:

A^TAΘ＝A^Tb.

suppose that A is^TA is conventional, and the parameters to be estimated are from:

Θ＝(A^TA)^-1A^Tb

and:

as mentioned above, the heating characteristic is characterized by two parameter sets, namely the slope θ_SAnd a horizontal theta_N. Therefore, the number of parameters q is 2. Thus, the result is:

abbreviated as:

L：＝A^TA，

r：＝A^Tb

this scheme can also be implemented succinctly, by:

Θ＝(A^TA)^-1A^Tb＝L^-1r

this representation then constitutes the starting point for the iterative estimation problem.

Weighted least squares method

The standard least squares method can be further generalized by weighting the individual rows of the data matrix and the weighting of the data points. For example, a scientific paper of the method will be presented in detail in this and the next sections. In Ljung, l.: system identification-user theory, preptic Hall, shangsader river (1999) and r.isermann and m.munchhof: identification of dynamic systems, schpringer (2010).

The weighted least squares method then forms a theoretical starting point for applying the least squares method to heating characteristic optimization. The weighting matrix can be expressed as follows

Wherein, w_iIs greater than 0 and

then, the weighted minimum multiplication problem is solved:

suppose A^TWA is conventional, the solution of this equation (1) is:

Θ＝(A^T·W·A)^-1A^T·W·b (2)

or with abbreviations:

L：＝A^TWA，

r：＝A^TWb

this compact representation is:

Θ＝(A^TWA)^-1A^TWb＝L^-1r

iterative least squares method

By both methods it turns out that all the first data points are disadvantageous and therefore the data matrices a and b, e.g. the data packet D, have to be determined over the entire period₀To D_NMeasurement of (2), data packet D_NDeciding on the end of the measurement period may require a lot of computational effort. For large amounts of data, a large amount of memory space and powerful computer functions must be provided. Instead, it is advantageous that the optimal control parameter set is already determined during the measurement. When new packets are available. The memory space required and the necessary computing power for this purpose may be small. This can be achieved by an iterative representation of the least squares problem.

For the iterative approach, the matrix is represented as follows:

for example, the divisionShould be interpreted in such a way that the variable with the identification k-1 represents the result of the optimal control parameters from past packets (packets up to the point in time k-1). The identification k thus points to the newly generated data packet D_kThe data packet is used to update the new control parameter Θ_k。

Thus, according to equation (2):

Θ_k＝(A_k-1 ^TW_k-1A_k-1+A_k ^TW_kA_k)^-1(A_k-1 ^TW_k-1b_k-1+A_kTW_kb_k)

have abbreviations:

L_k：＝A_k ^TW_kA_k

r_k：＝A_k ^TW_kb_k

further, equation (1) can be expressed succinctly as:

Θ_k＝(L_k-1+L_k)^-1(r_k-1+r_k)

for calculating Θ_kThe process of steps (a) can be simplified to use packet D as follows₀To D_NExamples of (2).

1. For data packet D₀In the first step,

L₀：＝A₀ ^TW₀A₀und r₀：＝A₀ ^TW₀b₀has already been calculated. This can be used to determine Θ₀＝(L₀)^-1(r₀)。

2. In a second step, the following is for the data packet D₁Is calculated by

L₁：＝A₁ ^TW₁A₁und r₁：＝A₁ ^TW₁b₁And theta₁＝(L₀+L₁)^-1(r₀+r₁) Is determined here.

3. Therefore, the following is for this packet D_kCalculation in the third step

L_k：＝A_k ^TW_kA_kAnd r_k：＝A_k ^TW_kb_k

From there the following can be calculated

Θ_k＝(L_k-1+L_k)^-1(r_k-1+r_k)

4. Then, L_k-1→L_k-1+L_kAnd r_k-1→r_k-1+r_kIs set and the operating variable k is increased by one.

5. If k is N +1, the procedure is terminated. Otherwise, return to step 3.

The following are examples of heating characteristic optimization:

for a daily calculation with a time resolution of 15 minutes, the parameter is N96; q is 2. Only the parameter theta needs to be set_k，L_k-1，r_k-1Stored in the storage medium 7 of the server 8 and thus compared to conventional methodsAt significantly less storage requirements. Basically, all parameter optimization problems of the control that can be represented by equation (1) can be managed and calculated efficiently on the server 8 using the procedure according to the invention described above.

Adapted to the heating characteristic curve

The following shows how the above method can be used to optimize the heating characteristics. Heating characteristics describe the external temperature T_ATarget indoor temperature T_R，WAnd a target supply temperature T_VL，WAnd is described by equation (3):

here, a_αAre known, previously determined parameters of the heating profile.

Equations (1) and (3) yield the data matrix:

having parameters to be determined

Rule of deviation

e_VL：＝T_VL，W-T_VL

e_R：＝T_R，W-T_R

Measured value T using a temperature sensor as shown in FIG. 2_VL(t)，T_A(t)，T_R(t)，T_VL，W(t)，T_R，W(t) are recorded at regular time intervals and sent as a time series in the form of data packets to the server 8 where they are stored and evaluated. Fig. 4 shows how the heating characteristic can be determined from the measured values. In FIG. 4, theThe measurements are plotted in a chart. The plot data points obtained from the measurement values may also be filtered using a weighting function such that only those deviations e are used_VLAnd e_RVery small data points. The heating characteristic curve fitted to the data points will then provide the optimum control parameter Θ_kIt may be transmitted to the heating controller 3.

For example, a gaussian distribution curve can be used as the weighting function:

parameter σ of the setting_VLAnd σ_RAre all essentially freely selectable. Depending on the number of data packets available for evaluation or the degree of dispersion of the measured values, the set parameters may be set larger or smaller to determine the optimum control parameter set Θ_k。

Second exemplary embodiment: system identification

The second exemplary embodiment relates to the temperature T of the reference room 9 in the building, compared to the heating characteristic curve of the first exemplary embodiment_RExplicit control of (2). The second exemplary embodiment describes how the internet and stored data packets are used to determine the dynamic model M. Model M is the supply temperature T_VLAnd an external temperature T_ATo room temperature T_RA mathematical description of the relationship between. The model is determined based on the measured data.

As described above, the data packet is generated and transmitted to the server 8 using the first exemplary embodiment. Therefore, only the control parameters Θ used for calculating the optimum are dealt with in the following_kDetails of the mathematical method of (1).

The mathematically constructed model allows the control parameters of the controller 3 of the HVACR system 2 to be specifically optimized according to the model, for example by means of P (proportionality), PI (proportionality, integral) or PID (proportionality, integral, differential) controllers. The optimization can be performed, for example, by standard control methods, such as root-trace curves, frequency-line methods, and the like. In comparison with the first exampleAs in the illustrative embodiment, the determination of the building model parameters and the optimization of the control parameters Θ_kIs executed on a server 8, which server 8 communicates with the control via an internet connection. The controller 3 itself works locally, for example in the heater of the heating system 2, and automatically resets the parameters at regular intervals (for example daily, weekly or monthly).

According to a second exemplary embodiment of the present invention, a model M of the controlled system is used to optimize the control parameters. For this purpose, the calculation algorithm executed on the server 8 is based on the measurement data obtained, i.e. the data packet D_kA mathematical relationship between the input and output data of the control process is determined.

Fig. 6a shows an initial situation of an exemplary process P (here a heating circuit). External temperature T_AIs a measurable disturbance variable. Supply temperature T of heating circuit_VLIs an input variable. Room temperature T_RIs an output or control variable. In addition, the unmeasurable disturbance variable is represented by G ═ G₁，...，g_q]And (4) showing.

The purpose of the method is to determine a mathematical model M such that the output variables provided by the model M are

Corresponds as precisely as possible to the measured room temperature T_R，As shown in fig. 6 b. Thus, the purpose is to make the output variable

With measured room temperature T_RThe deviation e between is minimal.

As a simple method of the model M, for example, the supply temperature G may be selected_VL(Z) time discrete transfer function in frequency domain and external temperature G_A(Z) a time-discrete transfer function therefor

T_R(Z)＝G_VL(Z)·T_VL(Z)+G_A(Z)·T_A(Z)

The time t is denoted below as the sampling time t_SA multiple of (b), thus: t is k.t_s. Time discrete transmission in frequency domainThe transfer function is here represented as a time-discrete transfer function in the Z-transformed image domain.

The transfer function in the frequency domain corresponds to a difference equation in the time domain

N_PHere the number of poles of the transfer function, and N_Z，VLOr N_Z，AIs the number of corresponding zero positions. In principle, these are freely selectable, but can generally be determined beforehand on the basis of physical considerations.

Mathematical model M uses an external temperature T_AAnd room temperature T_RThe historical values of (a) are calculated and the model is output, and the historical values are stored as data packets in the storage medium 7 of the server 8. Model parameter a_i，b_VL，i，b_A，IThe calculation of (c) can again be expressed as a least squares problem as in the first exemplary embodiment: a Θ is b. Matrices A and b may be derived from packet D_iAnd (4) generating. If other disturbance variables are known, e.g. solar radiation G_solOr an external heat source input P_FWThe model can be extended accordingly.

To compute the model parameters, the problem is converted to a least squares problem. Data packet D₁To D_NFor this purpose. For example, for a sampling time t of 15 minutes_sN96 is used for one day. So that:

k＝1，...，96

N_p＝2

N_Z，VL＝1

N_z，A＝1

from equation (4):

T_R(k)＝-a₁T_R(k-1)-a₀T_R(k-2)+b_VL，1T_VL(k-1)+b_VL，0T_VL(k-2)+b_A，1T_A(k-1)+b_A，0T_A(k-2)

thus, from D_1，...D_NTo obtain:

so that it can be transmitted from the data packet D to the server 8_iResulting in matrices a and b. Can be derived from the control parameter Θ ═ a₁，a₀，...，b_A，0]Calculating a transfer function G_VL(Z) and G_A(Z). If the transfer function G_VL(Z) and G_A(Z) is known, the controller 3 (e.g. P, P1 or PID controller) can be designed to control the room temperature T_R. This can be done using standard control techniques based on the model, such as root trajectory curves, frequency response curves, etc.

Since the parameters of the model M are updated periodically, for example daily, weekly or monthly, the result is an adaptive or optimal control of the heating system 2 (here the heating circuit) adapted to the process. The control parameters of the controller 3 are updated periodically accordingly. This allows continuous internet-based optimization of the controller 3.

The advantage and innovation of this method is that the identification can be made on the embedded system of the heat generator without maintenance and processing parameters, since these calculations are made on the corresponding capacities of the central platform on the internet.

The features disclosed in the foregoing description, in the claims and in the accompanying drawings may, both separately and in any combination thereof, be material for realizing various embodiments of the invention.

List of relevant labels

1. System for determining optimal control parameters

HVACR system

3. Closed-loop/open-loop controller

4. Heating loop controller

5. Boiler controller

6. Temporary memory (storage medium)

7. Database (storage medium)

8. Server

9. Room

10. Indoor temperature sensor

11. External temperature sensor

12. Supply temperature sensor

13. Mass flow sensor

14. Solar radiation sensor

15. Reflux temperature sensor

16. Heat radiator

17. Thermostat device

20. Interface for controller

21. Interface for server

22 device for generating data packets

CPU and processor

B. User' s