Method, apparatus and heat pump system for predictive regulation

Abstract

The invention relates to a method for predictive control of a heat pump system, in particular a data-driven predictive control method, in which, using a minimized defined objective function, at least one future input trajectory of the heat pump system is ascertained at least on the basis of system data of the heat pump system measured at an earlier time, the system data at least comprises input measurement data and output measurement data of the heat pump system. According to the invention, a predictive regulation of the heat pump system, in particular a data-driven predictive regulation, is limited to at least one first operating load range of at least one heat pump of the heat pump system and to at least one second operating load range of the heat pump of the heat pump system, the first operating load range and the second operating load range do not overlap each other and are spaced apart from each other.

Description

Technical Field

[0001] The invention relates to a method and a device for data-driven predictive regulation of a heat pump system. The invention also relates to a heat pump system. Background Art

[0002] A method for data-driven predictive regulation of a multi-energy system (including a cogeneration plant, a boiler and a thermal energy storage plant) is already known from the scientific publication "Towards data-driven predictive regulation of multi-energy distribution systems" by D. Bilgic et al. (Electric Power Systems Research, 212(2):108311), wherein at least one future input trajectory of the multi-energy system is determined using a defined objective function to be minimized, based at least on system data measured at an earlier time of the multi-energy system, wherein the system data at least include at least input measurement data and at least output measurement data of the multi-energy system. Summary of the Invention

[0003] The present invention is based on a method for predictive control of a heat pump system, in particular data-driven predictive control. In this method, at least one future input trajectory of the heat pump system, in particular at least a future input trajectory and a future output trajectory of the heat pump system, is determined based on at least system data of the heat pump system measured at an earlier time, using a defined objective function, in particular a defined objective function that is minimized according to defined operating limits. The system data include at least input measurement data and output measurement data of the heat pump system.

[0004] The present invention proposes to limit the predictive regulation of the heat pump system, in particular the data-driven predictive regulation, to at least one first operating load range of at least one heat pump of the heat pump system and to at least one second operating load range of the heat pump of the heat pump system, the first operating load range being in particular related to the input power and the second operating load range being in particular related to the input power, wherein the first operating load range and the second operating load range do not overlap with each other and are spaced apart from each other.

[0005] The method according to the present invention can advantageously improve the data-driven regulation of a heat pump system. Thus, regulation can be advantageously limited to an operating range within which the heat pump system can be operated technically and / or economically justified. The method according to the present invention can advantageously reduce parameterization overhead and / or improve the robustness of regulation. Since, in particular, impermissible operating ranges can be taken into account, better regulation performance can be advantageously achieved. Since, in particular, functionality and / or operation can be optimized and / or efficiency can be increased, costs can advantageously be reduced. Cost savings can also be advantageously achieved since computational overhead can be reduced.

[0006] In particular, in data-driven predictive control of system devices, physical modeling of the system devices is replaced by predicting the system behavior using system data (in particular, at least input and output measurement data). Thus, in data-driven predictive control, physical modeling of the actual system devices can be advantageously bypassed, with system behavior predicted solely using measurement data. A heat pump system is particularly configured to absorb thermal energy from a lower-temperature energy reservoir (e.g., ambient air, ground, etc.) while consuming technical work, and transfer this energy, along with drive energy (in particular, given by an input trajectory), as useful heat to a higher-temperature system to be heated (e.g., a building). To this end, the heat pump system particularly includes at least one heat pump, which may include various components, such as a compressor, a condenser, a throttle valve, and / or an evaporator. Furthermore, the heat pump system may also include a control and / or regulating unit and various heat exchangers, among others. Preferably, the input trajectory determined by the data-driven predictive control includes the electrical power supplied to the heat pump system, or the so-called power demand (power demand) of the heat pump system. In particular, in data-driven predictive control, in addition to determining the input trajectory, an output trajectory is also determined. It is also conceivable to determine another trajectory in data-driven predictive control, for example another input trajectory or another output trajectory. The output trajectory in particular includes the heat flow output by the heat pump system. The trajectory preferably includes a plurality of elements (Eintrag) of time steps that follow one another in time, preferably regularly. The time steps of the input trajectory preferably correspond to the time steps of the output trajectory. In predictive control, the "power request" (i.e., the electrical power demand) of the heat pump required to obtain a certain heat flow is preferably regulated.

[0007] In particular, input trajectory and output trajectory are obtained at least to a large extent based on the system representation of the heat pump system, which is based on the system data measured at an earlier time of the heat pump system. The system data of the heat pump system, in particular the system representation, at least includes a value pair of output measurement data and input measurement data of the heat pump system. The input measurement data can be constructed as an electric power input (operating load value / input load value) input to the heat pump system. For example, the input measurement data can be constructed as the electric power (or power demand) of the compressor flowing into the heat pump system and / or the forward flow temperature (Vorlauftemperatureatur) of the condenser side of the heat pump system. The output measurement data is especially constructed as heat flow measurement data corresponding to the input measurement data, which is for example on the condenser side of the heat pump system. Instead of value pairs, it is conceivable to incorporate other data (such as exogenous measurement data) to form value triplets, value quadruplets, etc. These value clusters (Wertebündel) preferably constitute system data. In particular, the system data is obtained at the moment before adjustment, for example, when manufacturing, calibrating or installing the heat pump system. It is conceivable that the system data can be updated, for example, within the context of maintenance, modification, repair, or recalibration. Furthermore, it is also conceivable that the heat pump system (particularly during operation) updates the system data autonomously, at least in part, at specific regular or irregular times. To this end, the heat pump system can have corresponding sensors. Furthermore, the heat pump system can (particularly autonomously or under the guidance of an operator) perform plausibility checks on the system data, thereby being able to identify system changes that may require an update of the system data.

[0008] Preferably, in data-driven predictive regulation, at least one future input trajectory is determined by minimizing an objective function (in particular a loss function, "cost function" in English). The objective function reflects, for example, technical and / or economic relationships. The objective function can minimize the electricity costs related to the time of day and / or the heat demand related to the time of day. For example, the objective function can be optimized by minimizing the energy costs of the operation while maintaining a minimum temperature (which may be related to the time of day). In particular, the goal of the optimization is to design the operation of the heat pump system so that, for example, energy costs and / or energy consumption are minimized. As a result, for example, the electricity demand during load peaks of national electricity demand can be reduced, so that the heat pump system advantageously contributes to improving grid stability. Alternatively or additionally, it is also possible to maximize efficiency and / or heating power and / or heating rate, etc. In data-driven predictive regulation, the objective function is minimized while taking into account the upper and lower operating limits defined for the heat pump system.

[0009] In particular, the operating load ranges to which the data-driven predictive regulation is limited are each constructed as a range of input power values ​​(power requests) that can be input for the heat pump of the heat pump system, which range is bounded by an upper limit and a lower limit. However, it is also conceivable that at least one operating load range is given by only one operating point, i.e., in particular for one operating load range, the upper limit may be identical to the lower limit. In particular, the upper limit of the first operating load range is less than the lower limit of the second operating load range. One of the operating load ranges in particular includes an operating load of zero. In particular, when the operating load is set to zero, the heat pump is in a no-load state. When the operating load is set to zero, the heat pump is disabled or enters a standby state (Standby-Zustand). In particular, the first operating load range does not have any operating load value / input power value that is also within the second operating load range. The operating load ranges in particular constitute a continuous operating load range. Preferably, the continuous operating load range does not include excluded operating load values / input load values. Preferably, the first operating load range and the second operating load range are respectively adjacent to an exclusion range, which is between the first operating load range and the second operating load range. Preferably, the upper limit of the first operating load range and the lower limit of the second operating load range usually delimit the exclusion range. The exclusion range constitutes in particular a range of operating load values / input load values ​​that should not / are not allowed to be set by regulation. At the operating load values / input load values ​​of the exclusion range, unstable operation of the heat pump of the heat pump system is expected. "Setting" should be understood in particular as specially programmed, designed and / or equipped. The object setting for determining a function should be understood in particular as: the object realizes and / or implements the determination function in at least one application state and / or operating state.

[0010] Furthermore, it is proposed that the first operating load range only includes zero operating load. This can advantageously incorporate the shutdown and / or standby operation of the heat pump into data-driven predictive regulation. In particular, "zero operating load" is equivalent to "no operating load". Preferably, the first operating load range only includes the shutdown operating state and / or the no-load (standby) operating state of the heat pump. The upper limit and the lower limit of the first operating load range preferably coincide. The second operating load range particularly includes the normal operating range of the heat pump, which extends, for example, between the full load of the heat pump and a non-zero minimum partial load, or between two partial loads. The normal operating range of the heat pump particularly includes the load range in which the heat pump can operate normally. The exclusion range particularly includes the load range in which the heat pump cannot operate normally. The normal operating range can, for example, consist of an operating load between 20% and full load of the heat pump. Of course, different lower limits (for example, 10% or 30% of full load) and / or different upper limits can also be considered.

[0011] Furthermore, it is proposed to additionally determine and preferably regulate the input trajectory of the heat pump system based on external measured data, such as the temperature of the heat source of the heat pump. This can advantageously further improve the functionality of the heat pump system, in particular data-driven predictive regulation, because in particular, external measured data that influences the performance of the heat pump system can be additionally taken into account. This can also advantageously increase efficiency. The external measured data can be, for example, the temperature on the evaporator side of the heat pump system. Preferably, the (particularly future) input trajectory determined by the regulation depends at least on the input measured data and / or the output measured data and / or the external measured data and / or other (external or internal) measured data deemed relevant by a person skilled in the art. Furthermore, it is conceivable to additionally determine and preferably regulate the input trajectory of the heat pump system based on other input measured data, which are also measured at an earlier time and are derived from interfaces with other parts / components of the heat pump system (for example, the inflow temperature on the condenser side of the heat pump). It is also conceivable to determine multiple input trajectories of the heat pump system simultaneously (but separately) in the data-driven predictive regulation. For example, the input measurement values ​​of the input measurement data, the output measurement values ​​of the output measurement data, the exogenous measurement values ​​of the exogenous measurement data, and the additional (exogenous) input measurement values ​​of other input measurement data together constitute a relevant measurement data group of the system data (here a value quad), and then data-driven predictive adjustments are performed based on this.

[0012] Furthermore, it is proposed that, for predictive regulation of a heat pump system, a control equation, in particular a system of control equations, populated with system data, be solved using a control and / or regulation unit of the heat pump system, or using a computing unit arranged externally to the heat pump system and in communication with the heat pump system. This advantageously enables efficient regulation of the heat pump system, in particular analytically solvable regulation. The control and / or regulation unit preferably includes a solver for solving the control equations. Alternatively, an external computing unit, in particular a central computing unit (e.g., a cloud), may include a solver for solving the control equations and / or have access to a solver for solving the control equations. The control equations may also constitute a system of control equations or be represented as a system of control equations. In particular, the control equations include at least one measurement data matrix and a decision variable vector on one side (left side) of the equation and a trajectory vector on the other side (right side). The control equations may preferably be represented as matrix equations. A "control and / or regulation unit" should be understood in particular to mean a unit having at least one control electronics unit. "Control electronics unit" should be understood in particular to mean a unit having a processor, a memory unit, and an operating program stored in the memory unit. A "solver" is to be understood as meaning, in particular, a special mathematical computer program which can solve mathematical equations and / or systems of equations and / or terms and / or similar tasks. In the data-driven predictive control of the heat pump system, preferably a plurality of time steps of at least the input trajectory, in particular a plurality of time steps of all trajectories in the trajectory vector, are optimized simultaneously. In the method for controlling the heat pump system, the trajectory vector is preferably determined and / or updated repeatedly with a fixed time step. In particular, at least the future input trajectory determined in the control constitutes a prediction of the heat pump system within the prediction range. The control equations constitute, in particular, equations, in particular a system of equations, for a data-driven system representation of a heat pump or heat pump system.

[0013] Furthermore, it is proposed that a control and / or regulating unit or a computing unit determine a trajectory of the heat pump system, in particular at least one input trajectory and at least one output trajectory of the heat pump system, using a control equation based on a matrix-vector multiplication of at least one measurement data matrix and a decision variable vector. The measurement data matrix is ​​formed from two or more stacked Hankel matrices, each of which includes only one measurement data type (e.g., only one input measurement data type, such as electrical power or condenser-side forward flow temperature; or only one output measurement data type, such as generated heat flow) and is modified by binary variables. This allows for optimized heat pump control, as rapid switching between normal operation above a minimum power and shutdown (in particular, at least into no-load operation) can be taken into account when determining the future input trajectory of the heat pump system, in particular when regulating the heat pump system. In particular, multiple measurement data types can also be included in the measurement data matrix for inputs, outputs, and (if necessary) external inputs. Accordingly, the trajectory vector on the right side of the control equation can also contain the corresponding trajectories of these multiple measurement data types. This advantageously allows for better control performance, since in particular impermissible operating ranges can be taken into account. Compared to control concepts that do not take this into account, improved performance and / or cost savings can be achieved.

[0014] The data-driven predictive system description known from the prior art already includes the control equation (1) with a control matrix that is different from the measurement data matrix and consists of a plurality of vertically stacked Hankel matrices with a known matrix notation. The Hankel matrix of the known control matrix can contain the input measurement data u d , on the other hand contains the output measurement data y d Specifically, the measurement data matrix includes a separate Hankel matrix for each measurement data type, which is arranged vertically within the measurement data matrix. In the known regulation equation, these Hankel matrices are interconnected via the decision variable vector g to determine a trajectory vector containing the input trajectory u and the output trajectory y. The input trajectory u and the output trajectory y each form a vector.

[0015]

[0016] The known control equations do not allow the exclusion of operating load ranges from the control. In addition, when using equation (1) to control heat pump systems with a strongly nonlinear operating behavior, it becomes more difficult to approximate the dynamics of the corresponding heat pump system. In order to overcome these drawbacks, the measurement data matrix according to the invention has at least two, or three or more modified Hankel matrices, which respectively include the input measurement data u d , output measurement data yd and / or possibly additional exogenous measurement data w d The modified Hankel matrices of the measurement data matrix are preferably stacked one on top of the other with known matrix notation. In the modified Hankel matrix, each matrix element is preferably expanded by a binary variable. The expansion includes, in particular, multiplying the corresponding matrix element by the binary variable.

[0017] The regulation equation according to the invention is represented below as equation (2). In order to perform data-driven predictive regulation of the heat pump system, the solver preferably solves the following equation (2).

[0018]

[0019] On the left side of equation (2) is a measurement data matrix and a decision variable vector, the measurement data matrix having two modified Hankel matrices stacked one above the other. If additional exogenous measurement data is considered, for example, another modified Hankel matrix containing exogenous measurement data can be inserted between the modified Hankel matrices shown in equation (2). On the right side of equation (2) is a trajectory vector having input trajectories and output trajectories. If exogenous measurement data is considered, the trajectory vector also includes an additional exogenous data vector. The modified Hankel matrix above contains the input measurement data The modified Hankel matrix below contains the output measurement data In equation (2), no additional exogenous measurement data are taken into account. In this case, the superscript d indicates that this relates to input measurement data of the heat pump measured at an earlier moment. The subscript i indicates to which relevant measurement data set determined within a common time step the individual measurement data elements belong. The subscript i extends from the value 1 (first time step of the first measurement data set / trajectory) to the value T (last time step of the trajectory). N represents the prediction horizon within which the regulation should be carried out. The prediction horizon indicates, in particular, how far into the future the situation should be taken into account in each iteration of the data-driven predictive regulation. The prediction horizon indicates the number of future input trajectory values ​​that should be determined in each iteration of the regulation. For example, in the case of hourly readjustment of the heat pump, the prediction horizon can assume the value N=12 or N=24. In this case, T ini represents the number of known initial values ​​that should be included in the regulation. It is conceivable that the regulation can be carried out without taking into account the initial values. Alternatively, it is also conceivable to include the initial values ​​(for example the corresponding current measured values ​​u ini or y ini ) into the adjustment. This case is exemplified in (2). However, it is also conceivable to include other past measurements immediately before the initial value into the adjustment. In this case, there is T ini>1. In this case, T corresponds to the length of the corresponding measurement data trajectory. The measurement data trajectory has, in particular, T elements. In particular, the vertical arrangement of the modified Hankel matrix within the measurement data matrix corresponds to the vertical arrangement of the trajectory vectors. The number of elements of the decision variable vector corresponds, in particular, to the number of columns of the measurement data matrix. The number of elements of the trajectory vector corresponds, in particular, to the number of rows of the measurement data matrix.

[0020] The decision variable vector contains elements g l , where the index l corresponds to the column of the measurement data matrix. Thus, the index l in particular represents the column number of the measurement data matrix. The index l can correspond to the index i of the aforementioned measurement data matrix element. The trajectory vector contains the elements In this case, the superscript x indicates the trajectory to which the element of the trajectory vector belongs. The input trajectory is denoted by u and the output trajectory is denoted by y. The subscript j is from 1 to N+T. ini Here, the elements It is also possible to enter the trajectory u k Or output trajectory y k The indices ini and 1 to N represent the initial time step "ini" (whose measured value has been determined) and the future time steps 1 to N to be evaluated in the regulation.

[0021] Binary variable δ k can only take the value 0 or 1 (see also formula (3)). By modifying the elements of the Hankel matrix of the measurement data matrix with the binary variable, the corresponding elements are either set to zero (δ k =0), or remain unchanged (δ k =1).

[0022]

[0023] Additionally, it is proposed that each element of a row of a control equation be assigned the same binary variable, which is in particular assigned to a time step of a data-driven predictive control. This allows for optimized heat pump control, in particular because entire rows of control equations can be completely opened or closed when solving the control equations (in particular, a system of control equations). Because impermissible operating ranges can be taken into account, improved control performance can advantageously be achieved. Compared to control schemes that do not take this into account, improved performance and / or cost savings can advantageously be achieved. Preferably, each matrix element of a measurement data matrix and each trajectory element of a trajectory vector in a row of the control equations is multiplied by the same binary variable. Each binary variable is in particular associated with a time step of the control prediction horizon. For example, if the binary variable for the third time step of the prediction horizon becomes 0, the row of the control equation associated with the third time step is completely closed, in particular set to 0=0. Preferably, all identical binary variables / all binary variables with the same index k in the control equations always have the same value. Preferably, all identical binary variables / all binary variables with the same index k in the control equations are always optimized together.

[0024] Furthermore, it is proposed that each element of each first row of each Hankel matrix in the modified Hankel matrix of the control equation, in particular each element of each first row of the trajectory vector, and preferably each element of each first row of the entire control equation, be assigned the same binary variable, which is in particular assigned to a time step for the data-driven predictive control. Advantageously, the control of the heat pump can be optimized, in particular because multiple relevant rows of the control equation, in particular the system of control equations, can be completely opened or closed when solving the control equation. Advantageously, improved control performance can be achieved, in particular because impermissible operating ranges can be taken into account. Advantageously, relevant rows of different measurement data types are always closed together via the binary variable. On the one hand, the introduction of the binary variable advantageously allows for rapid switching of the heat pump's operating state between a second operating load range representing the normal operating range and a complete shutdown (the first operating load range). On the other hand, the introduction of the binary variable in the output measurement data, in addition to the input measurement data, allows for a particularly accurate approximation of the heat pump's operating range (in particular, by eliminating dynamics during heat pump shutdown). This is particularly important for heat pumps with relatively strong nonlinear characteristics. Both aspects contribute to cost savings in heat pump operation. For each measurement data type (in the corresponding modified Hankel matrix on the left side of equation (2) and in the corresponding trajectory on the right side of equation (2)), the binary variables are renumbered so that the same time step for each measurement data type is always assigned the same binary variable. For example, the first row of each modified Hankel matrix and the first element of each trajectory of the trajectory vector receive the same binary variable δ1. The second row of each modified Hankel matrix and the second element of each trajectory of the trajectory vector receive the same binary variable δ2. And so on until the last time step N of the prediction horizon. The n-th row of each modified Hankel matrix and the n-th element of each trajectory in the trajectory vector preferably receive the same binary variable δ n . Preferably, identical time steps of each measurement data type have identical binary variables. In particular, the binary variable indices in the measurement data matrix are numbered in ascending order, wherein the binary variable index of the first row is set to 1. In particular, the index of a time step of the prediction horizon is identical to the index of the binary variable assigned to this time step. In particular, in the regulation equation (2), the index of a time step of the prediction horizon is identical to the index of the binary variable assigned to this time step. In particular, in the regulation equation (2), the index k of the binary variable assigned to a time step is different from the index i of the associated measurement data group representing the measurement data matrix.

[0025] It is proposed that for each prediction time step within the prediction range (see the right side of equation (2)), a binary variable (heat pump off: binary variable = 0; heat pump on: binary variable = 1) be introduced and multiplied. This has the advantage that the resulting heat pump input power (the result of the element of the input trajectory on the right side of equation (2) associated with this time step) is either within a defined operating range determined by the upper and lower limits of the second operating load range (when the binary variable = 1) or the input power is set to zero (when the binary variable = 0). Advantageously, low load ranges in which the heat pump should not / cannot be operated can thus be avoided while still being able to shut down the heat pump. By the proposed introduction and multiplication of a binary variable in each first row of the control equation, etc., the dynamics of the time steps that are shut down by means of the binary variable can be advantageously ignored when solving the control equation, and only the elements of the measurement data matrix in the remaining, non-shutdown (importantly relevant) operating range can be taken into account.

[0026] Furthermore, it is proposed that when solving the control equations, in particular when minimizing the objective function, by means of the control and / or regulation unit or the calculation unit, at least the binary variables of the modified Hankel matrix and in particular the decision variables of the decision variable vector are optimized. Advantageously, the solution of the control equations can be simplified to a mixed integer linear programming problem. In particular, when solving the control equations, in particular when minimizing the objective function, all binary variables are preferably optimized simultaneously by the solver. Furthermore, when solving the control equations, in particular when minimizing the objective function, the decision variables g of the decision variable vector can also be optimized. l .

[0027] It is further proposed that when solving the regulation equations with the aid of a control and / or regulation unit or a calculation unit, in particular when minimizing the objective function, at least the future trajectory elements of the heat pump system, in particular the future trajectory elements of at least one input trajectory and the future trajectory elements of at least one output trajectory, are optimized. Advantageously, the functionality of the heat pump system can be improved and / or costs can be saved. It is conceivable that the first element of the trajectory of the trajectory vector is measured and set (see "ini" in equation (2)). Thus, a fixed starting point can be incorporated into the regulation. It is also conceivable to incorporate more past measured values ​​by means of set elements in the trajectory vector. However, the use of a set measured first element in the trajectory vector is optional. The first element of each trajectory of the trajectory vector can also be a future trajectory element, which is determined in the regulation. In particular, if it is reasonable to assume that the heat pump system is a static system, the setting of the first element of the trajectory of the trajectory vector can be omitted.

[0028] Additionally, it is proposed that, in order to solve the regulating equation, in particular to minimize the objective function, in particular by a control and / or regulating unit or a computing unit, at least one row of the regulating equation is replaced by two inequalities (in particular inequalities of opposite signs). Advantageously, the computational overhead can be reduced, since the solution of the regulating equation can be simplified to a "mixed integer linear programming problem". By introducing inequalities, the problem to be solved, in particular by the solver, is advantageously relaxed. Preferably, each row of the regulating equation is replaced by two inequalities. In this way, all other rows are replaced by corresponding inequalities. In this case, the left side of the regulating equation is set once to be greater than or equal to the right side and once to be less than or equal to the right side. In this way, the mathematical meaning of each inequality group consisting of two inequalities corresponds to the mathematical meaning of the corresponding row of the original regulating equation.

[0029] In addition, it is proposed that, in order to solve the control equation, in particular by a control and / or regulating unit or a computing unit, each trajectory element of the two inequalities is expanded by a large M term (in particular by addition or subtraction), the large M term (Big-M-Term) in particular having different signs. Advantageously, the computational overhead can be reduced, since this serves in particular to simplify the solution of the control equation to a partial step of a "mixed integer linear programming problem". The large M term comprises at least a number M, which is designed as an arbitrarily large number. Preferably, the number M is much larger than the expected value of the trajectory. In particular, the value of the number M is at least large enough to enable the solver to solve the control equation reliably and / or quickly. For example, the value of the number M can be greater than 1,000,000 or greater than 1,000,000,000, or even greater.

[0030] Furthermore, it is proposed that, in order to solve the control equations, in particular by a control and / or regulating unit or a computing unit, binary variables are set for the large M terms of the two inequalities so that the large M terms are activated when the binary variable has the value 0 and are disabled when the binary variable has the value 1. Advantageously, the computational overhead can be reduced, since this is particularly used to simplify the solution of the control equations to a partial step of a "mixed integer linear programming problem". By advantageously converting the optimization problem into a linear representation, the solution of the optimization problem can be made more efficient for the solver, in particular because the problem becomes less complex. This means that either the required computing power can be reduced or, with unchanged computing power, a more precise control can be achieved (for example, by expanding the prediction range that can be considered with unchanged computing power). Advantageously, by introducing binary variables on the large M terms on the left side of the inequalities, the use of binary variables on the left side of the inequalities can be advantageously omitted. The inequalities expanded with large M terms (which together replace a row of the control equations) are shown in formula (4):

[0031]

[0032] In the first inequality (right side is greater than or equal to the left side), the large M term is added to the trajectory element. In the second inequality (right side is less than or equal to the left side), the large M term is subtracted from the trajectory element. The large M term is designed so that when the binary variable is 0, the value M is added to the first inequality and subtracted from the second inequality. The large M term is designed so that when the binary variable is 1, the value 0 is added to the first inequality and subtracted from the second inequality. Through this selection / design of the large M term, the binary variable being 0 allows any result to appear on the left side, in which case the left side can be less than a very large positive M and greater than a very large negative M. This is approximately equivalent to turning off the corresponding row of the control equation, because this row no longer affects the solution of the control equation, especially the minimization of the objective function. Through this selection / design of the large M term, the binary variable being 1 allows the inequality system to be simplified again to a single-row system of equations, because now only the equal sign can provide a valid solution to the inequality system. In this case, the corresponding rows of the control equations are still relevant / effective for solving the control equations, in particular for minimizing the objective function. In particular, the two inequalities do not contain binary variables on the left side (representing the left side of the control equations). In particular, the large M terms of the two inequalities each have a binary variable. However, the task of the binary variables (in particular, switching off individual rows) remains unchanged. The switching off of individual rows by the solver corresponds in particular to the solution result of the control, according to which the heat pump should be switched off at the corresponding time in the forecast horizon. In particular, when optimizing the binary variables, the solver determines the time step in the forecast horizon at which the heat pump should be switched off in order to achieve the optimal objective function.

[0033] Furthermore, it is proposed that, in order to solve the control equation, in particular to minimize the target function, in particular by a control and / or regulation unit or a calculation unit, preferably for the purpose of relaxing the control equation, the trajectory elements of the two inequalities, in particular modified by the binary variable, are replaced by parameters determined by two further inequalities. Advantageously, the problem to be solved by the solver can be completely linearized. Advantageously, the control can be improved and / or accelerated. In particular, the trajectory elements of the inequalities are replaced by the parameters Instead, the parameter is determined by the following inequality in formula (5):

[0034]

[0035] The following effect is produced in particular by the formulation of formula (5): when the binary variable δ k = 0, the trajectory element The value of is limited to 0 upward and downward; when the binary variable δ k =1, the trajectory element The value of is determined / limited by the set minimum and maximum limits (the lower and upper limits of the second operating load range).

[0036] Furthermore, it is proposed that a control and / or regulation unit or a calculation unit determine a trajectory of the heat pump system, in particular at least one input trajectory and at least one output trajectory of the heat pump system, by means of a regulation equation, by matrix-vector multiplication of at least two (preferably more than two) mutually related measurement data matrices with associated (in particular, preferably different) decision variable vectors that establish a correlation between the associated measurement data matrices. Each trajectory includes only a portion of all time steps of the prediction range, in particular only two consecutive time steps of the prediction range, wherein the two consecutive trajectories are each linked in such a way that the second row of the preceding trajectory constitutes the first row of the succeeding trajectory. Advantageously, regulation can be performed more accurately because the data description by the modified Hankel matrix can remain intact for further system representation and be taken into account in subsequent time steps. Advantageously, the associated measurement data matrices are preferably identical. In particular, the associated measurement data matrices each include only two rows of input measurement data and two rows of output measurement data. In particular, the trajectory vector associated with the respective measurement data matrix contains only two time steps, in particular two elements, for each measurement data type. In this case, the control equations that can be expressed as individual matrix equations are divided into a matrix equation system having a plurality of associated control equations, in particular a number corresponding to the number of time steps of the forecast horizon.

[0037] The following formula (6) describes the corresponding matrix equation system with the associated regulation equations:

[0038]

[0039] The aforementioned large-M method can also be applied to the implementation according to formula (7). Thus, the corresponding large-M terms can be modified as shown in formula (7) so that if the relevant binary variable is optimized to 0 in one of two time steps of the corresponding control equation, the entire control equation in the matrix equation system with multiple connected control equations is always disabled. The correspondingly modified large-M terms can be obtained from formula (7):

[0040]

[0041]

[0042] Furthermore, a device for (particularly data-driven) predictive control of a heat pump system is proposed. The device is configured to determine at least a future input trajectory, particularly a future input trajectory and a future output trajectory, of the heat pump system based at least on system data of the heat pump system measured at an earlier time, using a defined objective function to be minimized, particularly within certain limits. The system data includes at least input measurement data and output measurement data of the heat pump system. The device is configured to limit predictive control of the heat pump system, particularly data-driven predictive control, to at least a first operating load range of at least one heat pump of the heat pump system and at least a second operating load range of the heat pump of the heat pump system, wherein the first operating load range and the second operating load range do not overlap and are spaced apart from each other. The method of the present invention advantageously improves data-driven control of the heat pump system. Advantageously, control can be limited to an operating range within which the heat pump system can be operated technically and / or economically. Advantageously, the method of the present invention reduces parameterization complexity and / or increases control robustness. Advantageously, improved control performance can be achieved, particularly because impermissible operating ranges can be taken into account. Advantageously, costs can be reduced, in particular because the function and / or operation can be optimized and / or the efficiency can be increased. Advantageously, costs can be saved because the computing overhead can be reduced.

[0043] Furthermore, a heat pump system is proposed, which has a device for data-driven predictive regulation of the heat pump system. Advantageously, the functionality and / or efficiency of the heat pump system can be improved.

[0044] The method, apparatus, and heat pump system according to the present invention are not limited to the aforementioned applications and embodiments. To achieve the functionalities described herein, the method, apparatus, and heat pump system according to the present invention may, in particular, have a number of individual elements, components, units, and method steps that differs from the numbers mentioned herein. Furthermore, for value ranges specified in this disclosure, values ​​within the stated limits are also to be considered disclosed and can be used in any manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Further advantages can be derived from the following description of the drawings. The drawings illustrate an embodiment of the present invention. The drawings, the description, and the claims contain numerous combinations of features. Those skilled in the art will appropriately consider these features individually and combine them into other meaningful combinations.

[0046] Figure 1 A schematic diagram of a heat pump system is shown having an apparatus for data-driven predictive regulation of the heat pump system;

[0047] Figure 2A schematic flow chart illustrating a method for data-driven predictive regulation of a heat pump system;

[0048] Figure 3a shows a schematic input power-time graph of a heat pump system;

[0049] Figure 3b A schematic output power-time graph of a heat pump system is shown;

[0050] Figure 4a Schematic input measurement data versus time graph showing system data used for data-driven predictive regulation of a heat pump system;

[0051] Figure 4b a schematic exogenous measurement data versus time graph illustrating system data used for data-driven predictive regulation of a heat pump system; and

[0052] Figure 4c Schematic output measurement data versus time graph showing system data used for data-driven predictive regulation of a heat pump system. DETAILED DESCRIPTION

[0053] Figure 1 A schematic diagram of a heat pump system 10 is shown. The heat pump system 10 has a device for data-driven predictive regulation of the heat pump system 10. The device is configured to perform a method for data-driven predictive regulation of the heat pump system 10. The heat pump system 10 has a control and / or regulation unit 22. Alternatively or additionally, the heat pump system 10 may also have a computing unit (not shown) arranged outside the heat pump system 10, which is in communication connection with the heat pump system 10. The heat pump system 10 includes a heat pump 36. The heat pump 36 has a compressor, a condenser, a throttle valve and / or an evaporator and other components required for the operation of the heat pump 36. The heat pump 36 is configured as a closed refrigeration cycle. The heat pump 36 has a heat exchanger (not shown). The heat exchanger can be configured to transfer the heat generated in the closed refrigeration cycle to domestic water (e.g., a floor heating device). However, other applications of the heat pump 36 are also conceivable. The device is configured to determine a future input trajectory 28 (see FIG3 ) of the heat pump system 10 based on system data 30 measured at an earlier time of the heat pump system 10 using a defined objective function to be minimized. An exemplary objective function is given in the following formula (8):

[0054]

[0055] Figure 2A schematic flow chart of a method for data-driven predictive regulation of a heat pump system 10 is shown. In at least one method step 24 (which is not necessarily part of the regulation method), system data 30 of the heat pump system 10 are determined. This can be done, for example, during the manufacture of the heat pump system 10, during the calibration of the heat pump system 10, or during the recalibration of the heat pump system 10. The system data 30 include at least input measurement data 12 and output measurement data 14. In addition, the system data 30 can also include external measurement data 20. Figure 1 In the example shown, in method step 24 the input measurement data 12 of the heat pump 36 are determined (see also Figure 4a The input measurement data 12 include, for example, the power request of the heat pump 36. Furthermore, in method step 24, the output measurement data 14 of the heat pump 36 are determined (see Figure 4c The output measurement data 14 include, for example, the generated heat flow. Furthermore, in method step 24, external measurement data 20 are determined (see Figure 4b The exogenous measurement data 20 include, for example, the temperature on the condenser side of the heat pump 36. In this exemplary embodiment, the determined input measurement data 12, the determined output measurement data 14, and the exogenous measurement data 20 form a data triple. The measurement values ​​of the data triple with the same index are each determined simultaneously.

[0056] In at least one further method step 26, a future input trajectory 28 of the heat pump system 10 is determined based on the system data 30 of the heat pump system 10 measured at an earlier time in method step 24. In method step 26, the future input trajectory 28 is determined using the defined objective function in formula (8). In addition, in method step 26, a future output trajectory 38 and all other trajectories of the trajectory vector are determined using the defined objective function in formula (8). To this end, the objective function is minimized / optimized. The determination of the future input trajectory 28 of the heat pump system 10 performed in method step 26 is achieved by means of a data-driven predictive control of the heat pump system 10. The data-driven predictive control of the heat pump system 10 performed in method step 26 is limited to a first operating load range 16 and a second operating load range 18 of the heat pump 36. The first operating load range 16 and the second operating load range 18 do not overlap with each other and are spaced apart from each other. The first operating load range 16 only includes zero operating load of the heat pump 36 (off state of the heat pump 36). The second operating load range 18 includes a load range of the heat pump 36 in which normal operation is possible.

[0057] In method step 26, the solver of the control and / or regulation unit 22 solves the regulation equations filled with system data 30 for predictive regulation of the heat pump system 10. To this end, the solver minimizes the objective function in formula (8). The regulation equation is shown in formula (2) above. Alternatively, the solver of the control and / or regulation unit 22 can also solve multiple interrelated regulation equations for predictive regulation of the heat pump system 10, which are also filled with system data 30. The interrelated regulation equations are shown in formula (6) above. The regulation equation includes a measurement data matrix multiplied by a decision variable vector on one side of the equation and a trajectory vector with future trajectories 28, 38 on the other side of the equation. The regulation equation includes binary variables. In method step 26, the solver optimizes the binary variables of the regulation equation when solving the regulation equation. In method step 26, the solver optimizes the decision variables of the decision variable vector of the regulation equation when solving the regulation equation. In method step 26 , the solver optimizes future trajectory elements of the input trajectory 28 and of the output trajectory 38 when solving the control equations.

[0058] In substep 42 of method step 26, rows of the control equation are replaced by two inequalities to simplify the solution of the control equation, in particular to simplify the minimization of the target function. In substep 42, all rows of the control equation are replaced by the two inequalities. For example, the two inequalities shown in formula (4) replace the second row of the control equation of formula (4). In at least one further substep 44 of method step 26, in order to simplify the solution of the control equation, in particular to simplify the minimization of the target function, the corresponding trajectory elements of the two inequalities are expanded once by addition and once by subtraction, as shown in formula (4). The large M terms of the two inequalities are set to binary variables such that the large M terms are activated when the binary variable has a value of 0 and are deactivated when the binary variable has a value of 1. In at least one further substep 46 of method step 26, in order to simplify the solution of the control equation, in particular to simplify the minimization of the target function, the trajectory elements of the two inequalities modified by the binary variables are replaced by parameters determined by the two further inequalities. This substitution can be seen from the combination of equations (4) and (5) above.

[0059] In at least one further method step 48, the result of the first trajectory element of input trajectory 28 that can be assigned to a future time step of prediction horizon 40 is set on heat pump 36. In at least one further method step 50, the control equations are solved again using new initial values ​​(i.e., the values ​​set in method step 48).

[0060] Figure 3aAn input power-time diagram 52 of a heat pump system 10 is shown. The relative input power of the heat pump 36 is plotted on the horizontal axis 62 of the input power-time diagram 52. Time is plotted in evenly spaced time steps on the vertical axis 64 of the power-time diagram 52. These time steps correspond to the number of time steps of the trajectory 28, 38 of the trajectory vector of the control equation. The total number of time steps for the data-driven predictive control results in a prediction range 40. In particular, the number of time steps, the prediction range 40 and the intervals between the time steps can be freely selected. The input trajectory 28 is recorded in the input power-time diagram 52. A first operating load range 16 is marked in the input power-time diagram 52. The first operating load range 16 comprises only the zero line of the input power-time diagram 52. In Figure 3a In the example shown, the input trajectory 28 is in the first operating load range 16 (δ k =0). The second operating load range 18 is marked in the input power-time graph 52. The second operating load range 18 includes a range between a lower limit 54 and an upper limit 56 of the relative input power. The lower limit 54 of the second operating load range 18 is at a relative input power of 0.2 (i.e., 20% of full load). The upper limit 56 of the second operating load range 18 is at a relative input power of 1 (i.e., full load). Figure 3a In the example shown, the input trajectory 28 is in the second operating load range 18 (δ k = 1). First operating load range 16 and second operating load range 18 do not overlap. First operating load range 16 and second operating load range 18 are spaced apart from each other. An exclusion range 58 exists between first operating load range 16 and second operating load range 18. Exclusion range 58 forces the two operating load ranges 16, 18 to be spaced apart from each other. Exclusion range 58 specifies the range of relative input power within which heat pump 36 cannot operate reliably, safely, and / or economically. Input trajectory 28 determined using the method according to the present invention cannot have any values ​​within exclusion range 58. Lower limit 54 and upper limit 56 are selectable / variable.

[0061] Figure 3b A graph 60 of the output power of the heat pump system 10 is shown. The output power of the heat pump 36 is plotted on the horizontal axis 66 of the output power-time graph 60. Time is plotted in evenly spaced time steps on the vertical axis 68 of the power-time graph 52. These time steps correspond to the number of time steps of the trajectory vectors 28 and 38 of the control equation. The output trajectory 38 is recorded in the output power-time graph 60.

Claims

1. A method for predictive regulation, in particular data-driven predictive regulation, of a heat pump system (10), wherein: At least one future input trajectory (28) of the heat pump system (10) is determined using a defined objective function to be minimized, at least based on system data (30) of the heat pump system (10) measured at an earlier time, wherein the system data (30) at least include input measurement data (12) and output measurement data (14) of the heat pump system (10), characterized in that predictive regulation of the heat pump system (10), in particular data-driven predictive regulation, is restricted to at least one first operating load range (16) of at least one heat pump (36) of the heat pump system (10) and at least one second operating load range (18) of the heat pump (36) of the heat pump system (10), wherein the first operating load range (16) and the second operating load range (18) do not overlap with each other and are spaced apart from each other.

2. The method according to claim 1, characterized in that The first operating load range (16) includes only zero operating load.

3. The method according to claim 1 or 2, characterized in that In addition, an input trajectory (28) of the heat pump system (10) is determined and preferably regulated based on external measurement data (20), such as the temperature of a heat source of the heat pump (36).

4. The method according to any one of the preceding claims, characterized in that For predictive regulation of a heat pump system (10), regulation equations filled with system data (30) are solved by means of a control and / or regulation unit (22) of the heat pump system (10) or by means of a calculation unit arranged outside the heat pump system (10) and having a communication connection with the heat pump system (10).

5. The method according to claim 4, characterized in that The control and / or regulation unit (22) or the calculation unit determines a trajectory (28, 38) of the heat pump system (10), in particular at least one input trajectory (28) and at least one output trajectory (38) of the heat pump system (10), by means of a regulation equation, based on a matrix-vector multiplication of at least one measurement data matrix and a decision variable vector, wherein the measurement data matrix is ​​formed from at least two Hankel matrices stacked one above the other, each containing only one measurement data type and modified by binary variables.

6. The method according to claim 5, characterized in that Each element of a row of the control equation is respectively assigned the same binary variable, which is assigned in particular to a time step of the data-driven predictive control.

7. The method according to claim 6, characterized in that To each element of each first row of each Hankel matrix in the modified Hankel matrix of the control equations, the same binary variable is assigned, in particular, to a time step of the data-driven predictive control.

8. The method according to any one of claims 4 to 7, characterized in that When solving the control equations by means of the control and / or regulating unit (22) or by means of the computing unit, at least the binary variables of the modified Hankel matrix and in particular the decision variables of the decision variable vector are optimized.

9. The method according to any one of claims 4 to 8, characterized in that When solving the control equations with the aid of a control and / or regulating unit (22) or with the aid of a calculation unit, at least future trajectory elements of the heat pump system (10), in particular future trajectory elements of at least one input trajectory (28) and future trajectory elements of at least one output trajectory (38), are optimized.

10. The method according to any one of claims 4 to 9, characterized in that In order to solve the regulation equation, in particular by a control and / or regulating unit (22) or by a computing unit, at least one row of the regulation equation is replaced by two inequalities.

11. The method according to claim 10, characterized in that To solve the control equation, in particular by the control and / or regulating unit (22) or by the computing unit, the corresponding trajectory elements of the two inequalities are each expanded by a large M term, in particular by addition or subtraction.

12. The method according to claim 11, characterized in that In order to solve the control equations, in particular by a control and / or regulation unit (22) or by a calculation unit, binary variables are set for the large M terms of the two inequalities so that the large M terms are activated when the binary variable has the value 0 and are deactivated when the binary variable has the value 1.

13. The method according to claim 11 or 12, characterized in that To solve the control equation, in particular by a control and / or regulating unit (22) or a computing unit, the trajectory elements of the two inequalities, in particular modified by the binary variables, are each replaced by parameters determined by two further inequalities.

14. The method according to any one of claims 4 to 12, characterized in that Trajectories (28, 38) of a heat pump system (10), in particular at least one input trajectory (28) and at least one output trajectory (38) of the heat pump system (10), are determined by a control and / or regulation unit (22) or a calculation unit with the aid of regulation equations based on matrix-vector multiplications of at least two mutually related measurement data matrices with associated decision variable vectors, wherein each trajectory comprises only a portion of all time steps of a prediction horizon (40), in particular only two successive time steps of the prediction horizon (40), wherein two successive trajectories are respectively linked in such a way that the second row of the preceding trajectory forms the first row of the succeeding trajectory.

15. A device for predictive regulation of a heat pump system (10), in particular data-driven predictive regulation, the device being configured to predictively regulate the heat pump system, in particular by means of a method according to any one of the preceding claims, the device being configured to determine at least one future input trajectory (28) of the heat pump system (10) based on at least system data (30) of the heat pump system (10) measured at an earlier time, using a defined objective function to be minimized, wherein: The system data (30) comprises at least input measurement data (12) and output measurement data (14) of the heat pump system (10), characterized in that the device is configured to limit predictive regulation of the heat pump system (10), in particular data-driven predictive regulation, to at least one first operating load range (16) of at least one heat pump (36) of the heat pump system (10) and at least one second operating load range (18) of the heat pump (36) of the heat pump system (10), wherein the first operating load range (16) and the second operating load range (18) do not overlap with each other and are spaced apart from each other.

16. A heat pump system (10) comprising the device according to claim 14.

Images