EP1134508B1 - Air-conditioning control method for a weather dependent building or installation area - Google Patents

Abstract

The method involves forming a computer model of the building or system region, automatically calling up weather forecast data (12) including future solar radiation intensity and wind speed, computing control data (30) for an air conditioning system based on the model, the weather forecast data and stored load data profiles (26) and transmitting the control data to the air conditioning equipment.

Description

The invention relates to a method for controlling the climate in a weather-dependent building or plant area.

For the control of climatic data, such as temperature, pressure and humidity, in industrial plants, in which, for example, heat-sensitive products are processed, it is known to provide microprocessor-based controls. In these regulations, current climate data are compared with specified target values. Due to a deviation of the target size from the current climate data, ie, for example, a deviation of the desired temperature from the current temperature, control data is calculated. In the event of a temperature deviation, this will cause the heating or cooling system to be controlled accordingly to reach and maintain the desired temperature. Because the controller is based on a comparison between current climate data and given Target variables, ie desired climate data, based and the system to be controlled has capacitive properties, the predetermined target values are reached only after a reaction time (delay).

If, for example, the cooling of a building area is required, since the building area heats up due to solar radiation, further heating takes place due to the time delay until appreciable cooling occurs through the cooling system and the desired temperature is reached. To achieve the specified target size in the shortest possible response time, the cooling or heating systems must be designed to be large. This results in high investment costs. Even with very large cooling or heating systems, a certain reaction time always occurs. Furthermore, the use of cooling or heating systems, which are oversized for conventional operation, the energy consumption increases.

Modern cooling and heating systems have instead of the relatively low cost in the acquisition costs radiator Luftheiz- or air cooling systems provided in the floors or walls pipe systems. The provision of piping systems in the building mass has health advantages over convective air heating or air cooling. The radiant heat or cold is especially perceived as pleasant. In such pipe systems provided in the building system, the building mass, ie for example the wall or the floor of a building area, is cooled or heated. Since this is a large mass to be heated or cooled, the reaction times are up to a predetermined target value of the climatic data reached, extremely long. The building mass has a high thermal inertia depending on the material and construction. Today's control components require a manufacturer-specific software. A combination of components from different manufacturers is therefore only possible to a limited extent. This can lead to expensive systems having to be operated in parallel because they can not communicate with each other.

Today's control concepts are largely based on the fact that the hardware and software of the control system are integrated on site into the system or building to be controlled. This requires cost-intensive parameterization and optimization of the system on site by skilled personnel.

For the real plant or the building, an active technology - for example, a heating system - dimensioned, which is to ensure with the help of an adapted control that future climate data will be met in subsequent operation. Today, either stored in the past and / or currently detected measured values is used. The response of the system to weather conditions is described by transfer functions, such as heating characteristics. However, these can in principle only very roughly describe the thermodynamic behavior of the system, since it is not distinguished according to the influences of individual components of which the system is composed.

In order to describe the complexity of processes in larger buildings or facilities in more detail, learning and knowledge-based control systems, such as neural Nets used. However, here too, all real components are combined into a "black box", so that in the case of impermissible deviations between desired and actual states, the analysis of the calculation results does not point to individual components which are causally responsible for the errors.

From Mats Johansson: "Local weather forecasts control the HVAC system in buildings" CADDETT NEWSLETTER, No. 4, 1999, XP002199760, a climate control method is known in which control data is calculated based on the outside temperature. For this purpose, a future outdoor temperature is used, which is known by a weather forecast. In addition to the future outside temperature takes into account other effects such as solar radiation, wind speed and building use. Here, the building is considered as a uniform heat source.

DE 42 02 688 A1 discloses a method for improving the prediction of local weather data, i. Outside a building occurring temperatures described.

The object of the invention is to control the climate in a weather-dependent building or plant area such that future climate data reach and comply with a predetermined target size with the least possible time delay and high accuracy.

The object is achieved according to the invention by the features of patent claim 1.

According to the method of the invention, the building or plant area is modeled as a computer model, so that the thermal influence of, for example, walls, machines arranged in the building, light sources and the like. known is. According to the invention, the method for controlling the climate in a building or plant area automatically polls weather forecast data. By polling the weather forecast data, future weather data such as temperature, air pressure, humidity, wind speed, and solar radiation intensity are known for certain future time points. Furthermore, at least one load data profile is stored for calculating control data. In the load data profile known building or plant internal influences are stored. This is, for example, the time course from heat sources such as machinery, lighting and the like. and / or moisture development.

According to the invention, the method for controlling the climate in a building or plant area automatically polls weather forecast data. By polling the weather forecast data, future weather data such as temperature, air pressure, humidity, wind speed, and solar radiation intensity are known for certain future time points. Furthermore, at least one load data profile is stored for calculating control data. In the load data profile known building or plant internal influences are stored. These are, for example, the time course of heat sources, such as machines, lighting and the like. and / or moisture development.

With the help of the weather forecast data and the load data profile, control data are then calculated to achieve a predetermined target value of future climate data. The climate data are physical variables that are influenced by the weather, such as temperature, pressure, humidity, etc. With this control data, the air conditioning units associated with the building or plant area are controlled. By air conditioners, the internal condition, i. the climatic data of the building or the technical system, influenced. These are, for example, radiators, cooling surfaces, ventilation systems, etc. and in process plants, for example, to heat exchangers for decoupling heat and cold.

The calculated control data are transmitted to the air conditioners. The transmission can be done immediately after the calculation take place, so that the control data are stored in the air conditioners and trigger corresponding switching operations at the calculated times. It is also possible to transmit the control data to the air conditioners only at the appropriate times.

If, for example, the temperature in a building area is to be kept constant, the building area must be heated or cooled depending on the future weather conditions and the temperature changes caused by internal heat / cold sources. Is due to the weather forecast e.g. It is known that the outside temperature will change at a certain time such that the room temperature in the building area exceeds the desired setpoint, i. As the future target size increases, cooling of the room may begin at an earlier time when the room temperature is not yet elevated due to the increased outdoor temperature. In this case, according to the invention, the load data profile is taken into account, since this is e.g. can cause some balance or additional heating. In particular, when laid in the walls and floors cooling tubes can hereby be achieved that before the heating of the room by the increased outside temperature, the temperature of the walls and floors is already lowered slowly, without this, the room temperature changes noticeably. Without weather forecast data, it would only be possible to react at a time when the room temperature has already increased.

Due to the consideration of weather forecast data according to the invention, early achievement of predefined target variables of future climate data can take place at an early stage Control data are calculated and transmitted to the air conditioners. Future climate data can thus be achieved very precisely. For example, there are no strong temperature increases before reaching the desired temperature. This also applies to other climate data, such as humidity, air pressure and other weather-dependent climate data. Since the air conditioners can be controlled at a very early time, based on the time at which the specified target size of the future climate data is to be achieved, it is possible to provide significantly smaller heating and cooling systems. This results in a considerable saving of investment costs and operating costs. The method according to the invention is a forward-looking control method.

Preferably occurring loads are stored in the load data profile within the building or plant area. In such load data profiles are known, occurring at certain times occurring climatic changes influences stored. For example, herein can be the increase of internal loads by the switching on of light or in industrial plants by the commissioning of heat-emitting machines or the like. be saved. For example, an occupancy plan can be stored as a load data profile for the climate control of meeting rooms or lecture halls. In load data profiles also humidity changes, which are caused for example by the presence of persons, can also be stored. Due to stored load data profiles, known internal influences on the climate are already taken into account in the calculation of the control data for the air conditioners in advance.

Preferably, the load data profiles are corrected by comparison with climate data obtained in the past. There is thus an uninterrupted feedback. This is particularly advantageous when load data profiles at the beginning of operation of a heating or cooling system, which operates according to the inventive method, sometimes difficult to predict. The feedback provides automatic error correction of any errors contained in the original profile. The correction can be done by weighting the stored profile or by a concrete change of individual values in the profile.

In a particularly preferred embodiment of the method according to the invention, furthermore, local weather data are taken into account. By measuring local weather data, the weather forecast data can be corrected. This is particularly advantageous if only relatively inaccurate weather forecast data of weather services can be made available for a building or plant location. This is particularly the case in areas in which only for a large regional area a weather forecast is available, so that the weather forecast data locally inevitably have some inaccuracy.

Preferably, the weather forecast data is corrected by comparison with stored local weather data for the same or a comparable period of time. For example, if the weather forecast data for one day is the weather forecast data of the previous day, and a deviation between the local weather data and the weather forecast data is known from the local weather data of the previous day the weather forecast data for the current day are corrected immediately. Likewise, it is possible to form mean values from the local weather data measured over several days past and to use these as comparison data for the weather forecast data.

Preferably, the above method steps are repeated at regular short intervals of, for example, a few minutes. The query of weather forecast data is hereby not repeated as often as there are generally only at intervals of several hours or longer new weather forecast data. By repeating the method steps, local external influences, which are not taken into account in the determination of the control data for the air conditioners, can be compensated. This could be, for example, the opening of windows or the unforeseen switching on of the climate data influencing devices.

In a preferred embodiment, climate data profiles are provided. Climate data profiles store the required time course of the relevant climate data. This is, for example, a required temporal temperature profile as a function of the day of the week and the time of day. Based on the climate data profiles, the individual control data are calculated for specific times. In the climate data profiles, the target values for future climate data are thus stored in a time-dependent manner, so that these are target data.

The climate data profiles are preferably corrected on the basis of the climate data measured in the building or plant area.

Unpredictable influences, i. especially human behavior. By correcting the climate data profile, it can be taken into account, for example, that a person ventilates more frequently even on cold days than was assumed in the original climate data profile.

The measurement of the current climate data can be done continuously. In this case, when a predetermined limit value is exceeded, a new calculation of the optimal control data is automatically carried out.

The algorithm provided for the calculation of the control data is preferably constructed in such a way that a series of control data for temporally different future target variables are calculated in a calculation run. These are, for example, control data for the air conditioners for a period of several hours or even several days.

The calculation model by which the thermal influences of a building or installation area, ie, for example, walls, machines, light sources and the like arranged in the building, are recorded as well as possible is used to calculate the control data with regard to the thermal behavior. The building or the system is shown on a computer as a simulation model. This makes it possible to design the control system for the heating and air conditioning system at an early planning stage of the building. This has the advantage that the design of the heating and air conditioning and the associated investment costs can be defined at an early stage.

It is known to design and optimize systems or buildings that undergo weather-dependent, thermodynamic processes already in the planning and design phase by means of a computer-aided tool. In dynamically calculating simulation programs, mathematical-physical models are used that accurately describe the thermodynamic behavior of the system (plant or building). According to the invention, the system model is composed of modular, object-oriented software components which describe the behavior of a real component, for example a wall structure or a valve, more or less precisely. Such modules are stored in program libraries and can be used to compile any system. Furthermore, the required simulation models are given the required climate, availability and load data profiles, which determine the climate in the modeled system with the help of a likewise modeled control.

The influence of the weather is represented by the use of so-called test reference years, which resolve statistically representative weather situations, for example in the hourly space, for one year. The weather is therefore assumed to be known for a longer period of time. If the weather is applied to the system model, the simulation can be used to predict the influence of the weather on the thermodynamic behavior of the building or installation.

In contrast to the simulation, the real system today only has the knowledge about the weather of the past up to the current time of observation. The consequence is that the In principle, the climate in the system to be regulated can only react to the influence of the weather with a time delay. This delay is attempted to be compensated by large reserve capacities, such as a cooling system.

Instead, in the invention, the knowledge about the future weather development is integrated into the control strategy, thus the weather is assumed to be known for a longer future period and a control strategy optimally adapted to the real system is predicted. The system is thus able to not only react, but to actively act with foresight, with the result that the resulting climate in the system differs minimally or not at all from the required climate profile. The adjustment of the required climate is achieved with much smaller reserve capacities, because the system can be driven forward for a longer period of time rather than short-term with high performance with low reserve power.

To predict the optimal control strategy, the same models are used as in the simulation in the planning phase. In program libraries, modular, object-oriented software modules are stored, each describing the behavior of a real component, such as a wall structure or a valve, more or less precisely. Furthermore, the program library contains so-called functional modules, which perform infrastructural tasks, such as the measurement and archiving of weather data, condition data of the system to be controlled, automatic remote data transmission, among others.

All modules are clearly defined with regard to interface and data transmission protocol, so that they can be reused, interchanged and interlinked at any time. Although the modules can describe manufacturer-specific components, such as pumps or window constructions, they can be flexibly used in networking with one another to form a functional system and are not system-specific. Thus, using the program library, models for simulation, control and regulation of any system can be compiled.

In the planning phase, the real plant or building to be planned or already existing is depicted in detail as a simulation model in the computer. The created model is used both for the planning simulation and later for use within the regulation of the real plant. In the transition from planning to practical operation, the simulation model is supplemented by function modules. They represent the interface of the computer model with the system to be controlled and with data sources such as weather station, weather forecast, etc.

With the aid of the function modules, a logbook can be kept which provides information on the causes of the fault in the event of a fault. Furthermore, the knowledge of plant characteristics can be archived to be available for later planning and dimensioning tasks. This knowledge can be useful for scientific and business purposes. For fault and fault detection relevant states are detected by means of sensors in the system to be controlled by measurement. They are logged and compared with stored setpoints (profiles). Becomes a certain threshold value of the deviation between the setpoint and the actual state is exceeded, a warning or an automatic error correction is triggered within the control system.

The components of the program library work cross-platform, i. The software can be implemented in any hardware (microcontroller, PC etc.).

The calculation model used to control the climate data may be software or hardware, e.g. as a chip, be present. By mapping the building or plant area in the chip, the required computing power and thus the computing time can be reduced.

Preferably, a first buffer is provided in which the calculated control data is stored. As a result, the air conditioners can be supplied with control data even in the event of a failure of the computer over a longer period. A failure of the computer thus does not mean that the entire climate control fails.

Likewise, a second intermediate memory may be provided, in which the climate data and local weather data (state data) measured in the building or plant area are stored. Thus, a calculation can be carried out even in case of temporary failure of the measuring sensors in the building or plant area due to existing status data.

The inventive method can be separated by means of a single computer also for several building areas of a building or for several plant areas of a plant be used so that the climate in these areas can be controlled separately.

According to the invention modern means of communication, such as Internet, intranet, ISDN, telephone, etc., are used. They serve the networking of several subsystems and for remote maintenance and optimization. In this way, plant parameters and possibly models of individual plant components can be inexpensively remotely diagnosed and changed from an external data center instead of at the site of the plant.

The networking makes it possible to resort to distributed and / or central processing power.

Furthermore, the integration of the model-based control in already existing building and plant control technology is possible through the networking.

The control can be separated from the location of the system or building to be controlled. The investments and the on-site installation effort are reduced compared to conventional control technology. At the site of the plant remain only sensors, actuators and data buffer. If necessary, controls for setting desired profiles (climate, load data, availability of air conditioners) are also provided. By means of the operating elements it is possible for the plant operator, despite outsourcing the control, to independently make changes to nominal values. In a building, for example, the caretaker without special expertise, the desired room temperature in the Change the climate data profile or the planned room occupancy in the load data profile.

The actual calculation of the control strategy can be done externally by a service provider. He has computing power and the corresponding simulation models. The provider takes over the calculation of the optimal control strategy and sends the result automatically, for example via ISDN, back to the building. The provider can serve several properties in this way. He has the appropriate calculation models for each system to be controlled. Should changes be made to individual systems in the course of the operating time, for example in a building replacing a heating burner, the provider merely exchanges or supplements the corresponding module in the system model in order to adapt the control strategy to the new circumstances. In terms of a further reduction of on-site installation work and the associated costs, the setting of the desired profiles instead of on-site personnel can also be taken over by an external provider. For this purpose it suffices e.g. a janitor's order by e-mail, fax or telephone.

Unauthorized access from outside to the computer-aided control system is intercepted by suitable filters and security mechanisms at the interfaces of the remote data transmission.

In the above method, it is particularly advantageous that it is suitable both in the planning phase for the design and pre-test of the regulation of a plant and building technology as well as directly for practical use in a real system is. This avoids double generation of the computer model - once for the planning and once for the operation of the air conditioners. This results in a significant reduction in costs. Furthermore, errors are avoided by double input of data. In particular, the communication between the individual areas during construction and planning of the building or plant is improved because the same calculation model is accessed. It is optional whether the control strategy is determined at the location of the system to be controlled or externally at a remote location.

The invention will be explained in more detail below with reference to a preferred embodiment with reference to the drawing.

The drawing shows a schematic flow diagram of a preferred embodiment of the method according to the invention.

From a regional weather station 10 at predefined times, for example every twelve hours, weather forecast data is automatically requested and stored in a cache 12. The time interval between two queries of the weather forecast data is determined depending on the frequency of the updated weather forecast data provided by the regional weather station 10 and the specific constants, such as the thermal constant of a building or facility 18.

From a local weather station 14 current weather data are transmitted and stored in a buffer 16. The local weather station 14 is arranged in the area of the building or the installation 18 whose climate is to be controlled.

In the error correction weather forecast data 20, the weather forecast data is compared with present local weather data. In this case, a comparison with local weather data from the past takes place, which is a period comparable to the future period to be controlled. A comparable period is, for example, the previous day or another day for which a comparable weather forecast was made. However, it is also possible to extract a comparable period from several days of the past, which in total exceed the length of the considered future period to be controlled, by means of a statistical evaluation.

In the error correction, a comparison of the present weather forecast data with a corresponding reference period takes place. In this case, it is ascertained to what extent the weather forecast of the regional weather station 10 can be transferred to the local weather station 14 on the building or the installation 18 or to what extent it must be corrected. If necessary, a correction of the weather forecast data takes place. In this way, systematic deviations between regional weather forecast and actually locally occurring weather are detected and compensated.

The error correction 20 is performed whenever there is either a new weather forecast, or between weather predicted and error corrected in the past and weather measured locally locally, a deviation exceeding an allowable threshold occurs. After completion of the error correction 20, the calculation of the control data 30 is started via a start unit 28.

In the error correction of climate data 34, initially only a comparison is made between the required climate data profile 36 and the climate data 22 measured in the comparison period. If the difference between the demanded and the actual climate data profile exceeds a threshold value, then the originally required climate data profile 36 within the Error correction 34 calculates a modified climate data profile overrating the measured trend. If this process is completed, a re-calculation of the control data 30 with the modified climate data profile is started via the start unit 28, which counteract the measured trend via the air conditioners in the building or the system.

In the building 18, climate data, such as the air temperature, the humidity or the air pressure and possibly consumption data, such as the electrical power for art lighting or for the operation of machines, measured by sensors, stored and transmitted for error correction 24 and 34. The climate data is stored in a buffer memory 22.

In the error correction load data 24 stored load data profiles 26 are checked. The stored load data profiles are known influences on the climate. This applies, for example, switching on heat-generating devices at a certain time. By comparing the current climate data with the required climate data profile 32 and the load data profiles 26 current load data are calculated. After a statistical evaluation of such load data calculated and stored in the past up to the present time, the values supplied by stored load data profiles 26 become corrected. In this case, the stored load data profiles, ie the load data previously determined, for example, at the time of the first startup of the system or building, are compared with the calculated current load data. Due to the comparison, a weighting of the existing load data profile or a specific change of the stored load data profile takes place.

Based on the error correction of the weather forecast data 20 and the error correction of the climate data 34, the start unit 28 is started. The start unit 28 has control mechanisms that coordinate the time intervals of the recalculation of control data 30. In this way, for example, the fact is taken into account that a time delay occurs through the building until the excitation by the calculated control data 40 takes place via the measurement of the climate data 22 which then ensues. The start unit 28 causes the start of the calculation of the control data 30 after the internal control mechanism has been released.

The calculation of the control data takes place via a system or building model 36. The building model is the mapping of the building or installation area 18 to be controlled on a computer with the aid of a simulation program. With the help of the simulation program, a variety of different components, for example for walls, windows, radiators, lighting, machines, etc., are shown. From these individual modules, the plant or building area 18 can be emulated mathematically. It is thus possible an exact mapping of the building or plant area on the computer. Depending on the requirements of the accuracy of the control method, the model can also be simplified accordingly be. This reduces the number of arithmetic operations.

The illustrated building or plant area can also be constructed as hardware, such as a chip. This has the advantage that the computing speed is considerably increased.

The plant model 36, i. the control model underlying the control, corrected load data profiles 24 and corrected weather forecast data 20 are provided for calculating the control data. Furthermore, in the calculation of the control data 30 availability profiles 38 of the air conditioners are taken into account. In availability profiles 38 is stored, at which time certain air conditioners can be used. For example, heat pumps may only be used at certain times, as there are often contracts with power generators that prohibit the use of heat pumps at peak times. In such availability profiles 38 can be further deposited that certain devices, for example, work on a night-time basis and are therefore available only at certain times of the night. It is also possible to change the availability profiles 38 based on measured data.

With the help of a calculation algorithm 36 control data for air conditioners are calculated on the basis of the computer model. For example, in the illustrated computational model 36 are the room sizes in the building 18, the thermal constants of the individual walls, the thermal constants of objects or devices in the building. considered.

The calculated control data is transmitted to a buffer 40. At the time points also calculated, the control data are transmitted to the air conditioners provided in the building 18. The time at which control data is communicated to one of the air conditioners is prior to the time a target size of future climate data, i. For example, a desired in a room at a future time temperature is to be achieved.

In the case of a weather-dependent installation, the building blocks depicted in the calculation model may be e.g. to pump, store, pipe networks, valves, heat or cold sources, etc. act. Accordingly, the control itself is modeled.

Claims (14)

1. Air-conditioning control method for a weather-dependent building or installation area, comprising the steps of:

   - imaging the building or installation area as a computer model,

   - automatically reading weather forecast data (12), including future solar irradiation intensity and future wind velocity,

   - measuring weather date (16) to correct the weather forecast data (12), the current weather forecast data (12) being corrected by a comparison of local weather data (16) stored in the past and data forecast for the same or a comparable period,

   - calculating control data (30) for air-conditioning apparatus associated to the building or installation area based on the computer model, the weather forecast data (12) and stored load data profiles (26) in which the occurrence and the course of known heat sources occurring in the building or installation area, to obtain a predetermined target value, predetermined in the form of a future climate data profile, and

   - transmitting the control data to the air-conditioning apparatus.
2. The method of claim 1, wherein climate data (22) are measured in the building or installation area and the stored load data profiles (26) are corrected by a comparison with the measured climate data and/or the climate data required in the climate date profiles (32).
3. The method of claim 1 or 2, wherein the weather forecast data (12) are read in predetermined intervals, preferably 12 hours.
4. The method of claim 3, wherein the time interval is determined by the thermal time constant of the building or installation area.
5. The method of one of claims 1 - 4, wherein a climate data profile is provided in which the required time-dependent course of future climate data is stored.
6. The method of claim 5, wherein climate data (22) in the building or installation area are measured and the stored climate data profile (32) is corrected by comparison to the measured climate data.
7. The method of claim 5 or 6, wherein the control data (30) are calculated automatically when a change in the weather forecast data (12) or the climate data profile (32) exceeds a limit value.
8. The method of one of claims 1-7, wherein a series of control data (30) is calculated for chronologically different future target values.
9. The method of one of claims 1-8, wherein the computer model (36) is composed of individual modular model elements.
10. The method of claim 9, wherein the computer model (36) is provided as hardware and/or software.
11. The method of one of claims 1-10, wherein the control data (30) are stored in a temporary latch (40).
12. The method of one of claims 1-11, wherein the measured climate data are stored in a temporary latch (22).
13. The method of one of claims 1-12, wherein the climate of a plurality of building areas of a building or a plurality of installation areas of an installation are controlled separately by a computer.
14. The method of one of claims 1-13, wherein a plurality of buildings and/or installations are controlled by a computer via data telecommunication.

Images