Modelica language-based complex air conditioning system fault simulation method and system

Abstract

The invention discloses a complex air conditioning system fault simulation method and system based on a Modelica language, and relates to the technical field of air conditioners. Comprising the following steps: establishing a plurality of general component models of the air conditioning system based on a Modelica language, adding corresponding connectors to establish a plurality of key equipment models, and connecting the key equipment models to form a Modelica dynamic simulation model of the air conditioning system; the Modelica dynamic simulation model of the air conditioning system is set; the Modelica dynamic simulation model of the air conditioning system is configured, and model calibration and verification are carried out; and performing fault injection, and outputting and generating fault simulation output data of the whole set of air conditioning system at different levels of'component-equipment-system 'in a system level and a specific fault scene. Complex dynamic behaviors of an actual air conditioning system can be truly reflected, dynamic simulation of system fault characteristics is achieved, high-quality whole set of air conditioning system fault operation data in a system-level and specific fault scene is generated, and key data support is provided for research and development and performance evaluation of fault detection and diagnosis algorithms.

Description

Technical Field

[0001] This invention relates to the field of air conditioning technology, and specifically to a method and system for simulating faults in complex air conditioning systems based on the Modelica language. Background Technology

[0002] Heating, ventilation, and air conditioning (HVAC) systems are crucial for maintaining indoor comfort and air quality in modern buildings, and are also a major source of building energy consumption. With constantly changing outdoor climates and indoor demands, HVAC systems exhibit complex dynamic characteristics characterized by multi-device coupling and strong nonlinearity. These systems are prone to various failures during long-term operation, leading to decreased energy efficiency, deterioration of the indoor environment, equipment damage, and even safety hazards. Therefore, developing efficient fault detection and diagnosis (FDD) technologies for HVAC systems is of great significance.

[0003] The development and validation of current mainstream FDD methods (based on data-driven, model-driven, and hybrid methods) heavily rely on large amounts of high-quality, diverse fault operation data. However, obtaining fault data in practical engineering faces significant challenges:

[0004] (1) Actively injecting faults is costly and destructive, and is usually not feasible;

[0005] (2) Passive collection of natural faults is time-consuming and has problems such as insufficient monitoring parameters, uncontrollable fault degree and poor data quality, which makes it difficult to meet the training and verification needs of FDD models.

[0006] Using computer simulation technology to generate fault data is an effective way to solve the data shortage problem. However, current simplified simulation models are either limited to modeling single devices and cannot reflect system-level fault characteristics; or they are based on traditional simulation languages ​​and have significant shortcomings in simulating the dynamic coupling and nonlinear characteristics of complex systems, resulting in limited realism of the generated data. Summary of the Invention

[0007] To address the aforementioned technical challenges, this invention proposes a method and system for simulating complex air conditioning system faults based on the Modelica language. This method can realistically reflect the complex dynamic behavior of actual air conditioning systems, achieve dynamic simulation of system fault characteristics, and generate high-quality, complete sets of air conditioning system fault operation data at different levels (component-equipment-system) under specific fault scenarios. This provides crucial data support for the development and performance evaluation of the FDD algorithm.

[0008] According to a first aspect of the technical solution of the present invention, a method for simulating faults in a complex air conditioning system based on the Modelica language is provided, comprising the following steps:

[0009] S1, Air Conditioning System Model Establishment: Based on the Modelica language, establish multiple general component models of the air conditioning system, add corresponding connectors between the multiple general component models to establish multiple key equipment models, and connect the multiple key equipment models to form the Modelica dynamic simulation model of the air conditioning system.

[0010] S2, Parameter Acquisition and Setting: Acquire parameter information of each device in the actual air conditioning system, and set the Modelica dynamic simulation model of the air conditioning system based on the parameter information;

[0011] S3, Model Calibration and Verification: Configure the Modelica dynamic simulation model of the air conditioning system according to the parameter information of each device, and perform model calibration and verification by comparing the simulation output data with the actual monitoring data until the deviation between the two is less than or equal to the preset threshold.

[0012] S4, Fault Injection and Simulation: Based on fault scenario identification of faulty equipment and parameter information, modify the parameter values ​​of corresponding components in the Modelica dynamic simulation model of the air conditioning system to inject faults, and output a complete set of air conditioning system fault simulation output data at different levels of "component-equipment-system" under specific fault scenarios.

[0013] Furthermore, in S1, the multiple general component models specifically include a heat exchanger model, a compressor model, a throttle valve model, a fan model, and a surface cooler model.

[0014] Furthermore, in S1, the multiple key equipment models specifically include a chiller unit model, an air handling unit model, a fan coil unit model, a cooling tower model, a chilled pump model, and a cooling pump model.

[0015] Furthermore, in step S1, the multiple key equipment models are connected according to the physical topology and thermodynamic cycle relationship of the actual air conditioning system to construct a complete Modelica dynamic simulation model of the air conditioning system.

[0016] Furthermore, in S1, a heat exchanger model is built based on the Modelica language. The heat exchanger model is a shell-and-tube heat exchanger, and the heat flow and mass flow characteristics of the heat exchanger are calculated based on the liquid volume fraction. The heat transfer process, mass conservation, and energy conservation process of the fluid in the heat exchanger are represented by the following mathematical model:

[0017] Shell-side equations:

[0018]

[0019] In the formula: Q i α represents the heat transfer of unit i; α is the heat transfer coefficient. A represents the liquid volume fraction; total n is the total heat transfer area; cells T represents the number of discrete units. w T represents the wall temperature. v T represents the steam temperature. l For liquid temperature;

[0020]

[0021] Where: m vlc The total mass of the gas-liquid equilibrium fluid on the shell side; t is time; m in,shell The mass flow rate flowing into the shell side; m out,shell This is the mass flow rate exiting the shell side;

[0022]

[0023] In the formula: U shell H is the total internal energy of the shell-side fluid. in,shell H represents the enthalpy of the fluid flowing into the shell side. out,shell Q represents the enthalpy of the fluid flowing out of the shell side. shell This represents the total heat transferred through the wall to the shell-side fluid.

[0024] Pipe-side equations:

[0025] m in,tube =m out,tube (3-4)

[0026] Where: m in,tube The mass flow rate at the pipe-side inlet; m out,tube This represents the mass flow rate at the pipe-side outlet.

[0027]

[0028] In the formula: U tube H is the total internal energy of the fluid on the pipe side. in,tube H represents the enthalpy of the fluid flowing into the pipe. out,tube Q represents the enthalpy of the fluid flowing out of the pipe. tube This refers to the heat transferred from the pipe side to the wall.

[0029] Wall equation:

[0030]

[0031] In the formula: ρ wall c is the density of the wall material. wall V is the specific heat capacity of the wall material. wall Let be the volume of the wall material.

[0032] Furthermore, in S1, the compressor model is an efficiency-based compressor model built using the Modelica language. The compressor model defines the compression process through speed, displacement, volumetric efficiency, isentropic efficiency, and effective isentropic efficiency, and calculates the mass flow rate, power, and enthalpy change during the actual compression process.

[0033] Furthermore, the mathematical equations of the compressor model are as follows:

[0034] Volumetric efficiency takes into account volumetric losses during the compressor's suction process:

[0035]

[0036] In the formula: λ eff,comp For compressor volumetric efficiency; m dot V represents the actual mass flow rate. displacement n is the compressor displacement. comp ρ is the compressor speed; suction Density of the inhaled gas;

[0037] Deviation between the isentropic efficiency quantization compression process and the ideal isentropic process:

[0038]

[0039] In the formula: η isen,comp h represents the isentropic efficiency of the compressor. isentropic,discharge h is the isentropic enthalpy of exhaust. suction Enthalpy of inhaled gas; h discharge This is the actual exhaust enthalpy;

[0040] Effective isentropic efficiency takes into account the additional losses from the compressor's mechanical components:

[0041]

[0042] In the formula: η effisen,comp P represents the effective isentropic efficiency of the compressor. shaft,comp This refers to the compressor shaft power.

[0043] Furthermore, in S1, a throttle valve model is established using the Modelica language. The throttle valve model is based on Bernoulli's equation and uses the effective flow area to define the throttling characteristics. Pressure drop is generated by controlling the fluid flow area. The throttle valve model supports dynamic adjustment of the flow area through external input to achieve real-time control of the flow rate.

[0044] Furthermore, the mathematical equations for the throttle valve model are as follows:

[0045]

[0046] Where: m flowFor the mass of fluid passing through the throttle valve; A eff Effective circulation area; P input P is the pressure of the fluid before it enters the throttle valve. output ρ is the pressure of the fluid after passing through the throttle valve. input This is the density of the fluid when it enters the throttle valve.

[0047] Furthermore, in S1, a fan model is established using the Modelica language. The fan model controls the fluid state by setting a preset mass flow rate. The flow rate strictly follows the set value, ignoring system resistance, and the inlet and outlet pressure difference depends on the resistance characteristics of other components in the flow path.

[0048] Furthermore, in the aforementioned wind turbine model, based on the set mass flow rate, wind turbine efficiency, and actual pressure rise, the internal energy transfer of the wind turbine model is described by the following three power formulas:

[0049]

[0050] In the formula: P hyd,fan dp represents the hydraulic power of the fan. fan This represents the actual pressure rise of the blower; m flow,fan ρ is the mass flow rate of the fluid passing through the fan; ρ is the fluid density.

[0051]

[0052] In the formula: P shaft,fan For fan shaft power; η fan For fan efficiency;

[0053]

[0054] In the formula: P drive,fan For the fan drive power; η drive,fan For fan drive efficiency.

[0055] Furthermore, in S1, the surface cooler model is a surface cooler model based on the finite volume method established by Modelica language. The surface cooler model is based on the finite volume method, and the surface cooler is discretized into N elements along the flow equation. Each element contains dynamic equations for the gas side, liquid side and wall side. The equations follow the laws of conservation of mass, energy and momentum, and are coupled with the heat transfer equation.

[0056] Furthermore, the mathematical equations for the surface cooler model are as follows:

[0057] Gas-side equations:

[0058] m g =m g,i =m g,i-1 (3-14)

[0059] Where: m g For gas-side mass flow rate; m g,i m is the mass flow rate of the i-th unit on the gas side; g,i-1 The mass flow rate of the (i-1)th unit on the gas side;

[0060] m g ·(h g,i-1 -h g,i )+Q g,i =0 (3-15)

[0061] Where: h g,i h is the enthalpy of the gas exiting the i-th unit on the gas side. g,i-1 Q is the enthalpy of the gas exiting the (i-1)th unit on the gas side. g,i The heat exchange between the gas and the wall in the i-th unit;

[0062]

[0063] In the formula: Δp g,i f is the pressure drop of the i-th unit on the gas side; g,i L is the friction factor of the i-th unit on the gas side; i The length of each discrete unit in the flow direction; D h,g ρ is the hydraulic diameter on the gas side. g,i v is the gas density of the i-th unit; g,i Let ζ be the gas flow rate in the i-th unit; i This is the local drag coefficient;

[0064] Q g,i =α g,i ·A eff,g,i ·(T g,i -T w,ext,i (3-17)

[0065] In the formula: Q g,i α represents the heat exchange between the gas and the wall in the i-th unit; g,i A is the gas convective heat transfer coefficient; eff,g,i T represents the effective heat transfer area on the gas side. g,i T represents the gas temperature of the i-th unit; w,ext,i The temperature of the outer side of the wall of the i-th unit;

[0066] Liquid side equations:

[0067] m l,i =m l,i-1 (3-18)

[0068] Where: m l,i m is the mass flow rate of the i-th unit on the liquid side; l,i-1The mass flow rate of the (i-1)th unit on the liquid side;

[0069]

[0070] In the formula: ρ l,i V represents the liquid density of the i-th unit on the liquid side; l,i Let u be the volume of the i-th unit on the liquid side; l,i h is the internal energy of the i-th unit on the liquid side; l,i-1 h is the specific enthalpy of the liquid in the (i-1)th unit on the liquid side. l,i Q is the specific enthalpy of the liquid in the i-th unit on the liquid side; l,i The heat transferred from the wall to the liquid in the i-th unit;

[0071]

[0072] In the formula: Δp l,i f is the pressure drop of the i-th unit on the liquid side; l,i D is the friction factor of the i-th unit on the liquid side; h,l v is the hydraulic diameter on the liquid side. l,i Let be the liquid flow rate in the i-th unit;

[0073] Q l,i =α l,i ·A l,i ·(T w,int,i -T l,i (3-21)

[0074] In the formula: α l,i A is the liquid convective heat transfer coefficient; l,i T represents the heat transfer area on the liquid side. w,int,i T represents the temperature inside the wall of the i-th unit; l,i Let be the liquid temperature of the i-th unit;

[0075] Wall equation:

[0076]

[0077] Where: M w For wall quality; c p,w Specific heat capacity of the wall surface; T represents the rate of change of the wall temperature of the i-th unit over time. w,i Q is the wall temperature of the i-th unit; cond,i-1→i Q represents the axial heat conduction from the (i-1)th unit to the ith unit. cond,i→i+1 The heat conduction along the axis from the i-th unit to the (i+1)-th unit.

[0078] Furthermore, in S1, based on the working principle of the chiller unit and the connection relationship of each component, the chiller unit model is established based on the heat exchanger model, compressor model and throttle valve model.

[0079] Furthermore, in S1, the air handling unit model, fan coil unit model, and cooling tower model are all based on the general structure of "fan model + surface cooler model" and are established based on the fan model and surface cooler model.

[0080] Furthermore, in S1, both the cooling pump model and the refrigeration pump model are second-order pump models. Based on the similarity law and the quadratic characteristic curve, the pressure increment is calculated according to the ratio of the actual speed to the rated speed. The actual speed is input from the outside through the mechanical port.

[0081] Furthermore, the mathematical equations for the cooling pump and refrigeration pump models are as follows:

[0082] The formula for calculating pressure increment is as follows:

[0083]

[0084] In the formula: dp pump dp0 is the actual pressure rise of the pump; ρ is the pressure rise at rated zero flow; pump The actual density of the fluid flowing through the pump; ρ0 is the rated density; n pump n is the actual speed of the pump; n0 is the rated speed; V flow V is the actual volumetric flow rate of the pump; flow,0 Rated zero differential pressure flow rate;

[0085] The energy balance in the pump is calculated using transient methods, taking into account the increase in liquid temperature caused by power loss. The calculation of hydraulic power, shaft power and power loss is as shown in formulas (3-24) to (3-26).

[0086] P hyd,pump =dp pump ·V flow (3-24)

[0087] In the formula: P hyd,pump The hydraulic power of the pump;

[0088] P shaft,pump =P loss,pump +P hyd,pump (3-25)

[0089] In the formula: p loss,pump For the power loss of the pump; p shaft,pump This refers to the pump's shaft power.

[0090]

[0091] In the formula: η0 is the rated efficiency of the pump; This is the power loss correction factor.

[0092] Furthermore, in step S2, the parameter information of each device includes:

[0093] Heat exchanger (evaporator / condenser): heat transfer coefficient, initial liquid volume fraction, initial liquid temperature, initial wall temperature;

[0094] Compressor: speed, displacement, volumetric efficiency, isentropic efficiency, effective isentropic efficiency;

[0095] Throttling valve: effective flow area;

[0096] Chiller units: cooling water flow rate, chilled water flow rate;

[0097] Fan: mass flow rate, fan efficiency, drive efficiency;

[0098] Surface cooler: heat transfer coefficient, initial liquid temperature, initial wall temperature, initial gas pressure drop;

[0099] Refrigeration pump / cooling pump: speed, pressure increment at zero volume flow rate, initial liquid temperature, nominal temperature, volume flow rate at zero differential pressure, nominal efficiency;

[0100] Building parameters: room temperature, load data.

[0101] Furthermore, S3 specifically includes:

[0102] S31: Select a verification time period and obtain the values ​​of the parameter information of each device within the verification time period;

[0103] S32: Set the parameters of the Modelica dynamic simulation model of the air conditioning system according to the parameter values, and simulate the simulation output data of each device during the verification period.

[0104] S33: Compare the simulation output data with the actual monitoring data. If the deviation is less than or equal to the preset threshold, it means that the model can reflect the actual operation of the air conditioning system well and no adjustment is needed. If the deviation is greater than the preset threshold, the key parameters need to be readjusted until the deviation is less than the preset threshold.

[0105] Furthermore, the preset threshold is 15%.

[0106] Furthermore, S4 specifically includes:

[0107] S41: Determine the fault scenario: Determine the fault type and fault severity, wherein the target fault type includes single fault and multiple fault;

[0108] S42: Identify faulty equipment and parameter information: Based on the fault type, identify the faulty equipment and its corresponding parameter information;

[0109] S43: Fault Injection and Simulation: Based on the faulty equipment and its corresponding parameter information, set the parameter values ​​of the corresponding components in the Modelica dynamic simulation model of the air conditioning system, complete the injection of specific fault types and fault degrees, and simulate the faults of the air conditioning system.

[0110] S44: Output system fault data: Output complete set of air conditioning system fault simulation output data at different levels of "component-equipment-system" under specific fault scenarios.

[0111] Furthermore, in step S42, based on the fault type, identifying the faulty device and its corresponding parameter information specifically includes the following triples consisting of (fault type, faulty component, corresponding model parameter):

[0112] (Condenser scaling, condenser, condenser heat transfer coefficient)

[0113] (Evaporator scaling, evaporator, evaporator heat transfer coefficient)

[0114] (Compressor malfunction, compressor speed)

[0115] (Throttle valve stuck, throttle valve, effective flow area)

[0116] (Fan malfunction, fan, fan mass flow rate)

[0117] (Scale or blockage in the surface cooler, surface cooler, surface cooler heat transfer coefficient)

[0118] (Abnormal cooling water flow rate, chiller unit, cooling water mass flow rate)

[0119] (Abnormal chilled water flow rate, chiller unit, chilled water mass flow rate).

[0120] According to a second aspect of the technical solution of the present invention, a complex air conditioning system fault simulation system based on the Modelica language is provided, the system operating based on the method according to any one of the above aspects, wherein the system includes:

[0121] An air conditioning system model building unit is used to build multiple general component models of an air conditioning system based on the Modelica language, add corresponding connectors between the multiple general component models to build multiple key equipment models, and connect the multiple key equipment models to form a Modelica dynamic simulation model of the air conditioning system.

[0122] The parameter acquisition and setting unit is used to acquire parameter information of each device in the actual air conditioning system and set the Modelica dynamic simulation model of the air conditioning system based on the parameter information.

[0123] The model calibration and verification unit is used to configure the Modelica dynamic simulation model of the air conditioning system according to the parameter information of each device, and to perform model calibration and verification by comparing the simulation output data with the actual monitoring data until the deviation between the two is less than or equal to a preset threshold.

[0124] The fault injection and simulation unit is used to identify faulty equipment and parameter information based on fault scenarios, modify the parameter values ​​of corresponding components in the Modelica dynamic simulation model of the air conditioning system, inject faults, and output a complete set of air conditioning system fault simulation output data at different levels of "component-equipment-system" under specific fault scenarios.

[0125] The beneficial effects of this invention are:

[0126] (1) This invention establishes multiple general component models, then multiple air conditioning system equipment models, and finally connects them to form the entire air conditioning system model, thus constructing a hierarchical, high-precision Modelica dynamic simulation model of a complex air conditioning system from bottom to top, consisting of "component-equipment-system". This model covers the component level, equipment level, and system level, and can modify the model at any level to realize the fault simulation of any component and the output of system-level fault characteristics, overcoming the shortcomings of existing simplified simulation models that are limited to a single device and cannot perform component-level fault simulation and system-level fault output.

[0127] (2) This invention establishes a dynamic simulation model of an air conditioning system using the Modelica language. Based on the object-oriented, equation-based, and multi-domain unified modeling characteristics of the Modelica language, the model can more realistically reflect the complex dynamic behavior of the air conditioning system, achieve high-precision modeling and high-reliability simulation results for complex systems, overcome the difficulties of modeling complex systems using traditional simulation languages, and the obvious shortcomings in simulating the dynamic coupling and nonlinear characteristics of complex systems, resulting in limited data fidelity.

[0128] (3) This invention is based on a bottom-up hierarchical modeling model of "component-equipment-system". By modifying the parameter settings of each level of the "component-equipment-system" model in the Modelica dynamic simulation model of the air conditioning system, it can flexibly and controllably simulate fault scenarios of different severity and fault types, covering system-level fault characteristics. It solves the problems of actively injecting faults (high cost and destructive) and passively collecting natural faults (long cycle, uncontrollable fault degree, and poor data quality) in actual engineering. Moreover, compared with the fault simulation of a single device, this method can better reflect the propagation and impact of faults in complex coupled systems. Specifically, by modifying the model parameters, various faults can be simulated flexibly and controllably, covering system-level fault characteristics, and generating high-quality complete set of air conditioning system fault operation data at different levels of "component-equipment-system" under specific fault scenarios. The main reason why this invention can achieve the above effects is that the model is a bottom-up, hierarchical air conditioning system model built from "component level" to "equipment level" to "system level" based on the Modelica language. Because this model is built from the bottom up, covering "component level," "equipment level," and "system level," its input parameters include those of component models, equipment models, and system models, and its output parameters also cover all levels of "component-equipment-system." However, traditional simulation models only model a single device (such as a chiller unit), and their input and output parameters only involve the input parameters of the established device, unable to handle component-level or system-level parameter input and output. Therefore, compared to traditional models, the model of this invention can simulate the failure of any component or device and output system-level failure operating parameters by modifying the parameter settings of the models at each level of "component-equipment-system." Taking chiller unit failure as an example, the chiller unit of this invention is composed of four main components: compressor, expansion valve, evaporator, and condenser. By changing the parameters of components such as the compressor and expansion valve, simulations of compressor failures and expansion valve failures can be achieved. However, if a chiller unit model is directly built, only the chiller unit parameters can be set, and the failure simulation of components such as the compressor and expansion valve cannot be achieved.

[0129] (4) The present invention simulates faults based on a hierarchical air conditioning system model established from the bottom up. It can generate a complete set of air conditioning system fault operation data at different levels of "component-equipment-system" under system-level and specific fault scenarios. It provides key and scarce data resources for the research, training, verification and performance evaluation of HVAC fault detection and diagnosis algorithms, and solves the data shortage problem faced in this field. Attached Figure Description

[0130] Figure 1 A flowchart of a complex air conditioning system fault simulation method based on the Modelica language according to the technical solution of the present invention is shown.

[0131] Figure 2 A comparative diagram is shown between the bottom-up modeling method according to the present invention and the existing direct modeling method; wherein, the left side is the bottom-up modeling method and the right side is the direct modeling method;

[0132] Figure 3 This diagram illustrates the structure of an air conditioning system for a commercial building according to an embodiment of the technical solution of the present invention.

[0133] Figure 4 A schematic diagram of a complex air conditioning system model of a commercial building is shown according to an embodiment of the technical solution of the present invention. Detailed Implementation

[0134] The technical solutions of the embodiments of this application will be clearly and thoroughly described below with reference to the accompanying drawings. The described embodiments are merely some embodiments of this application.

[0135] This invention proposes a method and system for simulating faults in complex air conditioning systems based on the Modelica language. Modelica is an object-oriented, non-causal, equation-based, multi-domain unified modeling language that uses non-causal mathematical descriptions and equations to represent system behavior, giving it unique advantages in dynamic modeling of complex air conditioning systems in real buildings.

[0136] like Figure 1 As shown, the main steps are as follows:

[0137] (1) Establishment of the air conditioning system model: The actual air conditioning system is simplified, and a hierarchical, high-precision complex air conditioning system model is established based on the Modelica language. The specific steps are as follows:

[0138] First, establish general component models of the air conditioning system, including heat exchanger models, compressor models, expansion valve models, fan models, and surface cooler models. Second, add corresponding connectors between the component models to establish key equipment models, including chiller unit models, air handling unit models, fan coil unit models, cooling tower models, chilled water pump models, and cooling pump models. Finally, connect the equipment according to the actual system characteristics to form a dynamic simulation model of the air conditioning system.

[0139] (2) Parameter Acquisition and Setting: Acquire parameter information of each component in the actual air conditioning system, and set the Modelica dynamic simulation model of the air conditioning system based on the acquired information. Key parameters include:

[0140] ① Heat exchanger (evaporator / condenser): heat transfer coefficient, initial liquid volume fraction, initial liquid temperature, initial wall temperature;

[0141] ②Compressor: Speed, displacement, volumetric efficiency, isentropic efficiency, effective isentropic efficiency;

[0142] ③ Throttling valve: Effective flow area;

[0143] ④ Chiller units: Cooling water flow rate, chilled water flow rate;

[0144] ⑤ Fan: mass flow rate, fan efficiency, drive efficiency;

[0145] ⑥ Surface cooler: heat transfer coefficient, initial liquid temperature, initial wall temperature, initial gas pressure drop;

[0146] ⑦ Refrigeration pump / cooling pump: speed, pressure increment at zero volume flow rate, initial liquid temperature, nominal temperature, volume flow rate at zero differential pressure, nominal efficiency;

[0147] ⑧ Building parameters: room temperature and load data.

[0148] (3) Model calibration and verification: First, select a verification time period and obtain the operating parameters of each component of the air conditioning system during the time period; second, set the operating parameters of the Modelica dynamic simulation model of the air conditioning system based on the obtained information, and simulate the operating data of the air conditioning system during the time period; finally, compare the simulated parameters with the monitoring parameters of the actual system. If the deviation is less than or equal to 15%, it means that the model can reflect the operation of the actual air conditioning system well. If the deviation is greater than 15%, the key parameters need to be readjusted until the deviation is less than 15%.

[0149] (4) Fault Injection and Simulation: Based on fault scenarios, faulty components and parameters are identified, and the parameter values ​​of corresponding components in the Modelica dynamic simulation model of the air conditioning system are modified to inject faults. This enables fault simulation of different severity levels and types, generating high-quality, complete sets of air conditioning system fault operation data at different levels (component-equipment-system) under specific fault scenarios. The specific steps are as follows:

[0150] (1) Determine the fault scenario: Determine the target fault type (single fault / multiple faults) and the fault severity;

[0151] (2) Identify faulty components and parameters: Based on the fault type, identify the faulty components and their corresponding model parameters. Specifically, injectable faults include, but are not limited to, the following triples consisting of (fault type, faulty component, corresponding model parameter):

[0152] (Condenser scaling, condenser, condenser heat transfer coefficient)

[0153] (Evaporator scaling, evaporator, evaporator heat transfer coefficient)

[0154] (Compressor malfunction, compressor speed)

[0155] (Throttle valve stuck, throttle valve, effective flow area)

[0156] (Fan malfunction, fan, fan mass flow rate)

[0157] (Scale or blockage in the surface cooler, surface cooler, surface cooler heat transfer coefficient)

[0158] (Abnormal cooling water flow rate, chiller unit, cooling water mass flow rate)

[0159] (Abnormal chilled water flow rate, chiller unit, chilled water mass flow rate).

[0160] (3) Perform fault injection and simulation: Based on steps (1) and (2), modify the parameter values ​​of the corresponding components in the Modelica dynamic simulation model of the air conditioning system, complete the injection of specific fault types and fault degrees, and perform fault simulation of the air conditioning system.

[0161] (4) Output system fault data: Output high-quality "component-equipment-system" fault operation data of the complete air conditioning system at different levels under specific fault scenarios.

[0162] The following section provides a more detailed explanation of each correction.

[0163] ●Establishment of the air conditioning system model

[0164] The actual air conditioning system is simplified, and a hierarchical, high-precision complex air conditioning system model is built based on the Modelica language. It is important to note that the "complex air conditioning model" involved in this invention is a model of multiple types of combined centralized air conditioning systems in actual large public buildings. The model fully covers the two typical cold source types: centrifugal chillers and screw chillers, and adopts a terminal architecture combining all-air systems (centralized) and air-water systems (semi-centralized). The conversion between typical centralized / semi-centralized / combined air conditioning system forms can be achieved by starting and stopping the equipment model. Based on the bottom-up, hierarchical modeling method of "component-equipment-system" in this invention, suitable general components and equipment modules can be selected according to the actual building air conditioning system form to flexibly construct typical air conditioning system models in different large public buildings.

[0165] The specific steps are as follows:

[0166] First, establish a general component model.

[0167] (1) Heat exchanger model

[0168] Since condensers and evaporators operate on the same principle, a similar heat exchanger model was built using the Modelica programming language. This model is a shell-and-tube heat exchanger, primarily calculating the heat flow and mass flow characteristics based on the liquid volume fraction. During the simulation, the liquid volume fraction can dynamically change, affecting the heat transfer efficiency and temperature parameters of the heat exchanger. The model strictly adheres to the three fundamental principles of mass conservation, momentum conservation, and energy conservation, and its mathematical model can accurately characterize the heat transfer, phase change, and dynamic conservation relationships of mass and energy during the heat exchange process.

[0169] Shell-side equations:

[0170]

[0171] In the formula: Q i α represents the heat transfer of unit i; α is the heat transfer coefficient. A represents the liquid volume fraction; total n is the total heat transfer area; cells T represents the number of discrete units. w T represents the wall temperature. v T represents the steam temperature. l For liquid temperature;

[0172]

[0173] Where: m vlc The total mass of the gas-liquid equilibrium fluid on the shell side; t is time; m in,shell The mass flow rate flowing into the shell side; m out,shell This is the mass flow rate exiting the shell side;

[0174]

[0175] In the formula: U shell H is the total internal energy of the shell-side fluid. in,shell H represents the enthalpy of the fluid flowing into the shell side. out,shell Q represents the enthalpy of the fluid flowing out of the shell side. shell This represents the total heat transferred through the wall to the shell-side fluid.

[0176] Pipe-side equations:

[0177] m in,tube =m out,tube (3-4)

[0178] Where: m in,tube The mass flow rate at the pipe-side inlet; m out,tube This represents the mass flow rate at the pipe-side outlet.

[0179]

[0180] In the formula: U tube H is the total internal energy of the fluid on the pipe side. in,tube H represents the enthalpy of the fluid flowing into the pipe. out,tube Q represents the enthalpy of the fluid flowing out of the pipe. tube This refers to the heat transferred from the pipe side to the wall.

[0181] Wall equation:

[0182]

[0183] In the formula: ρ wall c is the density of the wall material. wall V is the specific heat capacity of the wall material. wall Let be the volume of the wall material.

[0184] (2) Compressor Model

[0185] An efficiency-based compressor model is built using the Modelica language. This model defines the compression process by speed, displacement, and three types of efficiency (volumetric efficiency, isentropic efficiency, and effective isentropic efficiency), and calculates the mass flow rate, power, and enthalpy change during the actual compression process. Equations (3-7) to (3-9) are the core calculation equations of the model, used to dynamically simulate the performance characteristics of the compressor.

[0186] Volumetric efficiency takes into account volumetric losses during the compressor's suction process:

[0187]

[0188] In the formula: λ eff,comp For compressor volumetric efficiency; m dot V represents the actual mass flow rate. displacement n is the compressor displacement. comp ρ is the compressor speed; suction Density of the inhaled gas;

[0189] Deviation between the isentropic efficiency quantization compression process and the ideal isentropic process:

[0190]

[0191] In the formula: η isen,comp h represents the isentropic efficiency of the compressor. isentropic,discharge h is the isentropic enthalpy of exhaust. suction Enthalpy of inhaled gas; h discharge This is the actual exhaust enthalpy;

[0192] Effective isentropic efficiency takes into account the additional losses from the compressor's mechanical components:

[0193]

[0194] In the formula: ηeffisen,comp P represents the effective isentropic efficiency of the compressor. shaft,comp This refers to the compressor shaft power.

[0195] (3) Throttling valve model

[0196] A throttling valve model was built using the Modelica programming language. Based on Bernoulli's equation, the model defines the throttling characteristics using the effective flow area and generates pressure drop by controlling the fluid flow area. The model supports dynamic adjustment of the flow area via external input, enabling real-time flow control. The real-time formula for calculating mass flow rate is as follows.

[0197]

[0198] Where: m flow For the mass of fluid passing through the throttle valve; A eff Effective circulation area; P input P is the pressure of the fluid before it enters the throttle valve. output ρ is the pressure of the fluid after passing through the throttle valve. input This is the density of the fluid when it enters the throttle valve.

[0199] (4) Wind turbine model

[0200] A fan model is built using the Modelica language, which controls the fluid state by setting a mass flow rate. The flow rate strictly follows the set value, ignoring system resistance, and the inlet and outlet pressure difference depends on the resistance characteristics of other components in the flow path. Based on the set mass flow rate, fan efficiency, and actual pressure rise, the internal energy transfer of the fan model can be described by three power formulas (3-11) to (3-13).

[0201]

[0202] In the formula: P hyd,fan dp represents the hydraulic power of the fan. fan This represents the actual pressure rise of the blower; m flow,fan ρ is the mass flow rate of the fluid passing through the fan; ρ is the fluid density.

[0203]

[0204] In the formula: P shaft,fan For fan shaft power; η fan For fan efficiency;

[0205]

[0206] In the formula: P drive,fan For the fan drive power; η drive,fan For fan drive efficiency.

[0207] (5) Surface cooler model

[0208] A surface cooler model based on the finite volume method is established using the Modelica language. This model discretizes the surface cooler into N elements along the flow equations. Each element contains dynamic equations for the gas side, liquid side, and wall side. The equations follow the laws of conservation of mass, energy, and momentum and are coupled with heat transfer equations, which can accurately characterize the dynamic conservation relationships of heat transfer, mass, and energy in the fluid heat exchange process.

[0209] Gas-side equations:

[0210] m g =m g,i =m g,i-1 (3-14)

[0211] Where: m g For gas-side mass flow rate; m g,i m is the mass flow rate of the i-th unit on the gas side; g,i-1 The mass flow rate of the (i-1)th unit on the gas side;

[0212] m g ·(h g,i-1 -h g,i )+Q g,i =0 (3-15)

[0213] Where: h g,i h is the enthalpy of the gas exiting the i-th unit on the gas side. g,i-1 Q is the enthalpy of the gas exiting the (i-1)th unit on the gas side. g,i The heat exchange between the gas and the wall in the i-th unit;

[0214]

[0215] In the formula: Δp g,i f is the pressure drop of the i-th unit on the gas side; g,i L is the friction factor of the i-th unit on the gas side; i The length of each discrete unit in the flow direction; D h,g ρ is the hydraulic diameter on the gas side. g,i v is the gas density of the i-th unit; g,i Let ζ be the gas flow rate in the i-th unit; i This is the local drag coefficient;

[0216] Q g,i =α g,i ·A eff,g,i ·(T g,i -T w,ext,i (3-17)

[0217] In the formula: Q g,iα represents the heat exchange between the gas and the wall in the i-th unit; g,i A is the gas convective heat transfer coefficient; eff,g,i T represents the effective heat transfer area on the gas side. g,i T represents the gas temperature of the i-th unit; w,ext,i The temperature of the outer side of the wall of the i-th unit;

[0218] Liquid side equations:

[0219] m l,i =m l,i-1 (3-18)

[0220] Where: m l,i m is the mass flow rate of the i-th unit on the liquid side; l,i-1 The mass flow rate of the (i-1)th unit on the liquid side;

[0221]

[0222] In the formula: ρ l,i V represents the liquid density of the i-th unit on the liquid side; l,i Let u be the volume of the i-th unit on the liquid side; l,i h is the internal energy of the i-th unit on the liquid side; l,i-1 h is the specific enthalpy of the liquid in the (i-1)th unit on the liquid side. l,i Q is the specific enthalpy of the liquid in the i-th unit on the liquid side; l,i The heat transferred from the wall to the liquid in the i-th unit;

[0223]

[0224] In the formula: Δp l,i f is the pressure drop of the i-th unit on the liquid side; l,i D is the friction factor of the i-th unit on the liquid side; h,l v is the hydraulic diameter on the liquid side. l,i Let be the liquid flow rate in the i-th unit;

[0225] Q l,i =α l,i ·A l,i ·(T w,int,i -T l,i (3-21)

[0226] In the formula: α l,i A is the liquid convective heat transfer coefficient; l,i T represents the heat transfer area on the liquid side. w,int,i T represents the temperature inside the wall of the i-th unit; l,i Let be the liquid temperature of the i-th unit;

[0227] Wall equation:

[0228]

[0229] Where: M w For wall quality; c p,w Specific heat capacity of the wall surface; T represents the rate of change of the wall temperature of the i-th unit over time. w,i Q is the wall temperature of the i-th unit; cond,i-1→i Q represents the axial heat conduction from the (i-1)th unit to the ith unit. cond,i→i+1 The heat conduction along the axis from the i-th unit to the (i+1)-th unit.

[0230] Secondly, based on the general component model, a key equipment model is established according to the connection relationship between the components.

[0231] (1) Chiller unit

[0232] A chiller unit consists of four basic components: an evaporator, a condenser, a compressor, and a throttling valve, which are connected in series to form a closed system in which the refrigerant circulates.

[0233] According to thermodynamic principles, the ideal refrigeration cycle is based on the reverse Carnot cycle. The refrigerant completes the cycle through adiabatic compression, isothermal heat release, adiabatic expansion, and isothermal heat absorption, transferring heat from a low-temperature heat source to a high-temperature heat source. Due to various limitations, the actual refrigeration cycle consists of two isobaric processes, one adiabatic compression process, and one adiabatic throttling process. In the evaporator, the refrigerant absorbs heat from the object being cooled (water) and vaporizes into steam. The compressor continuously extracts the generated steam from the evaporator and compresses it. The high-temperature, high-pressure steam is sent to the condenser, where it releases heat to the cooling medium and condenses into a high-pressure liquid. After being depressurized by the throttling mechanism, it enters the evaporator again, vaporizes once more, and absorbs heat from the object being cooled, thus completing the cycle repeatedly.

[0234] This invention adopts a bottom-up modeling approach. Based on the working principle of the chiller unit and the connection relationship of each component, it establishes a Modelica simulation model of the chiller unit based on a general heat exchanger model, compressor model, and throttle valve model. Compared with direct modeling of a single chiller unit, it allows modification of the parameters of the heat exchanger, compressor, and throttle valve models, enabling component-level fault injection and simulation of heat exchanger scaling, compressor failure, and throttle valve jamming.

[0235] This invention employs a bottom-up modeling approach, allowing for a significantly wider range of configurable parameters compared to directly modeled models. Taking a chiller unit as an example, the bottom-up chiller unit of this invention is constructed by connecting heat exchangers (evaporators and condensers), expansion valves, and compressors. The configurable parameters of this model are as described in S2 for the heat exchangers, compressor, expansion valves, and chiller unit. Previous studies, which used three performance curves for direct modeling of chiller units, only allowed setting parameters for chilled water temperature, cooling water temperature, cooling capacity, and polynomial coefficients. Figure 2 As shown in the diagram, this is a comparison between bottom-up modeling and direct modeling. Specifically, in the bottom-up modeling on the left, each component has a mathematical model, while in the direct modeling on the right, only the equipment has a mathematical model. The direct modeling method on the left can only modify the chilled water and cooling water temperatures, cooling capacity, and polynomial coefficients. Compared to the bottom-up model, the direct modeling model has fewer parameters that can be set.

[0236] (2) Air handling units / fan coil units / cooling towers

[0237] Air handling units, fan coil units, and cooling towers all operate on the principle of heat exchange, transferring heat through the contact between fluid and heat exchange surfaces. Air handling units and fan coil units utilize fans to force air circulation and directly regulate the supply air temperature via the coils; while cooling towers enhance airflow and lower water temperature through fans. The core mechanism of all three relies on the synergistic effect of forced convection and the heat exchange medium to ultimately achieve temperature control. Therefore, in modeling, all three can be simplified to a general structure of "fan + surface cooler." Based on a general fan model and surface cooler model, Modelica simulation models of air handling units, fan coil units, and cooling towers can be established.

[0238] (3) Cooling pump / refrigeration pump

[0239] The cooling pump and refrigeration pump models are created using the Modelica language. Both are second-order pump models. The model is based on the similarity law and quadratic characteristic curve. The pressure increment is calculated based on the ratio of the actual speed to the rated speed. The actual speed is input from the outside via the mechanical port.

[0240] The formula for calculating pressure increment is as follows:

[0241]

[0242] In the formula: dp pump dp0 is the actual pressure rise of the pump; ρ is the pressure rise at rated zero flow; pump The actual density of the fluid flowing through the pump; ρ0 is the rated density; n pump n is the actual speed of the pump; n0 is the rated speed; V flow V is the actual volumetric flow rate of the pump; flow,0Rated zero differential pressure flow rate;

[0243] The energy balance in the pump is calculated using transient methods, taking into account the increase in liquid temperature caused by power loss. The calculation of hydraulic power, shaft power and power loss is as shown in formulas (3-24) to (3-26).

[0244] P hyd,pump =dp pump ·V flow (3-24)

[0245] In the formula: P hyd,pump The hydraulic power of the pump;

[0246] P shaft,pump =P loss,pump +P hyd,pump (3-25)

[0247] In the formula: p loss,pump For the power loss of the pump; p shaft,pump This refers to the pump's shaft power.

[0248]

[0249] In the formula: η0 is the rated efficiency of the pump; This is the power loss correction factor.

[0250] Finally, the various devices are connected according to the system characteristics to form a dynamic simulation model of a complex building air conditioning system based on the Modelica language.

[0251] ●Parameter Acquisition and Setting

[0252] Obtain parameter information for each component in the actual air conditioning system, and set the Modelica dynamic simulation model of the air conditioning system based on the obtained information. Key parameters include:

[0253] ① Heat exchanger (evaporator / condenser): heat transfer coefficient, initial liquid volume fraction, initial liquid temperature, initial wall temperature;

[0254] ②Compressor: Speed, displacement, volumetric efficiency, isentropic efficiency, effective isentropic efficiency;

[0255] ③ Throttling valve: Effective flow area;

[0256] ④ Chiller units: Cooling water flow rate, chilled water flow rate;

[0257] ⑤ Fan: mass flow rate, fan efficiency, drive efficiency;

[0258] ⑥ Surface cooler: heat transfer coefficient, initial liquid temperature, initial wall temperature, initial gas pressure drop;

[0259] ⑦ Refrigeration pump / cooling pump: speed, pressure increment at zero volume flow rate, initial liquid temperature, nominal temperature, volume flow rate at zero differential pressure, nominal efficiency;

[0260] ⑧ Building parameters: room temperature and load data.

[0261] ●Model calibration and verification

[0262] First, select a verification time period and obtain the operating parameters of each component of the air conditioning system during this period. Second, based on the obtained information, set the operating parameters of the Modelica dynamic simulation model of the air conditioning system and simulate the operating data of the air conditioning system during this time period. Finally, compare the simulated monitoring parameters with the actual system monitoring parameters. If the deviation is less than or equal to 15%, it indicates that the model can reflect the actual operation of the air conditioning system well. If the deviation is greater than 15%, the key parameters need to be readjusted until the deviation is less than 15%.

[0263] ● Fault Injection and Simulation

[0264] Based on fault scenario identification of faulty components and parameters, the parameter values ​​of corresponding components in the Modelica dynamic simulation model of the air conditioning system are modified to inject faults, thereby achieving fault simulation of different severity levels and types. This generates high-quality, complete set of air conditioning system fault operation data at different levels (component-equipment-system) under specific fault scenarios. The specific steps are as follows:

[0265] (1) Determine the fault scenario: Determine the target fault type (single fault / multiple faults) and the fault severity;

[0266] (2) Identify faulty components and parameters: Based on the fault type, identify the faulty components and their corresponding model parameters. Specific injectable faults include, but are not limited to:

[0267] Table 1. Types of Air Conditioning System Faults and Corresponding Parameters

[0268]

[0269] (3) Perform fault injection and simulation: Based on steps (1) and (2), modify the parameter values ​​of the corresponding components in the Modelica dynamic simulation model of the air conditioning system, complete the injection of specific fault types and fault degrees, and perform fault simulation of the air conditioning system.

[0270] (4) Output system fault data: Output high-quality "component-equipment-system" fault operation data of the complete air conditioning system at different levels under specific fault scenarios.

[0271] Example

[0272] The development of air conditioning system fault detection and diagnosis technologies urgently requires the support of a large amount of high-quality fault data. However, in actual engineering projects, fault data acquisition faces problems such as high cost, long cycle, and poor data quality. Existing simulation methods are limited to single-device modeling or the inherent limitations of traditional simulation languages, making it difficult to accurately simulate system-level fault characteristics. Based on this, this embodiment proposes a complex air conditioning system fault simulation method and system based on the Modelica language. By constructing a dynamic simulation model of a complex air conditioning system in a real building using Modelica, and injecting faults, dynamic simulation of system fault characteristics is achieved. This generates a complete set of air conditioning system fault operation data at different levels (component-equipment-system) with controllable fault severity, providing a key data foundation for the development and performance evaluation of fault detection and diagnosis technologies.

[0273] The example analyzed is the air conditioning system of a commercial building, such as... Figure 3 As shown, the building's air conditioning system employs a combination of centralized and semi-centralized systems. The chiller room is equipped with 7 chiller units, 9 chilled water pumps, and 9 cooling pumps, serving the entire building. Large spaces such as department stores, corridors, ice rinks, and theaters utilize a full-air system (i.e., centralized system), while smaller spaces such as specialty stores, brand stores, and private dining / entertainment rooms employ fan coil units plus a fresh air system (i.e., semi-centralized system, also known as an air-water system).

[0274] The proposed method comprises four steps: First, a dynamic simulation model of a complex air conditioning system is constructed based on Modelica; second, parameter information of each component of the actual air conditioning system is obtained, and the Modelica dynamic simulation model of the air conditioning system is set based on the obtained information; third, the operating parameters of the Modelica dynamic simulation model of the air conditioning system are set and simulated, the simulated parameters are compared with the monitoring parameters of the actual system, and the model is calibrated and verified based on the comparison results; finally, the parameters of each component in the Modelica dynamic simulation model of the air conditioning system are modified respectively to realize fault injection and simulation of different severity and fault types, generating system-level, fault-controlled, and high-quality fault operation data of the entire system.

[0275] The four steps described above are implemented as follows:

[0276] (1) Establishment of the air conditioning system model

[0277] The actual building air conditioning system is simplified, and a hierarchical, high-precision model of the complex actual building air conditioning system is built based on the Modelica language. The specific steps are as follows:

[0278] First, establish a general component model for the air conditioning system, specifically including the condenser model, evaporator model, compressor model, expansion valve model, fan model, and surface cooler model;

[0279] Secondly, add corresponding connectors to each component model to establish key equipment models, specifically including 7 chiller units, 9 chilled pumps, 9 cooling pumps, 7 cooling towers, 4 air handling units, and 4 fan coil units.

[0280] Finally, the various devices were connected according to the actual system characteristics to form a dynamic simulation model of the air conditioning system. The refrigeration system in this model is divided into two systems: a large system and a small system. The large system includes 5 screw chillers, 6 cooling pumps, 6 chilled water pumps, and 5 large cooling towers. The small system includes 2 centrifugal chillers, 3 cooling pumps, 3 chilled water pumps, and 2 small cooling towers. The specific model construction is as follows... Figure 4 As shown.

[0281] (2) Parameter acquisition and setting

[0282] The parameter information of each component in the air conditioning system of this commercial building was obtained through actual measurement, and the Modelica dynamic simulation model of the air conditioning system was set based on the obtained information. Taking a small system as an example, the key model parameter settings are as follows:

[0283] ① Heat exchangers (evaporator / condenser): The evaporator has an initial liquid volume fraction of 0.7, and both the initial liquid temperature and wall temperature are 20℃; the condenser has an initial liquid volume fraction of 0.8, and both the initial liquid temperature and wall temperature are 32℃; the shell-side heat transfer coefficient and the tube-side heat transfer coefficient of both are 2000 W / (K·m). 2 ).

[0284] ② Compressor: Speed ​​set at 70Hz, displacement at 0.0136m³ 2 The volumetric efficiency and isentropic efficiency are both 0.7, and the effective isentropic efficiency is 0.63.

[0285] ③ Throttling valve: Effective flow area is 2.715 cm² 2 .

[0286] ④ Chiller unit: chilled water mass flow rate is 329.17 kg / s, cooling water mass flow rate is 130.55 kg / s.

[0287] ⑤ Fans: The fan efficiency in the cooling tower is set to 0.4, the drive efficiency is 1, and the fixed mass flow rate is 1123.88 kg / s; the fan efficiency in the air handling unit and fan coil unit is set to 0.4, the drive efficiency is 1, and the mass flow rate is controlled by PI.

[0288] ⑥ Surface Coolers: The initial liquid temperatures of the surface coolers in the cooling tower, air handling unit, and fan coil unit are 12℃, 12℃, and 20℃, respectively; the initial wall temperatures are 25℃, 25℃, and 20℃, respectively; the initial gas pressure drop is 0 Pa; and the gas-side heat transfer coefficient is 100 W / (K·m). 2 ).

[0289] ⑦ Refrigeration Pump / Cooling Pump: The refrigeration pump speed is set to 50Hz, the pressure increment at zero volumetric flow rate is 4 bar, the initial liquid temperature is 30℃, the nominal temperature is 20℃, and the volumetric flow rate at zero differential pressure is 470 m³ / h. 3 / h, nominal efficiency of 0.6; cooling pump speed of 50Hz, pressure increment at zero volumetric flow rate of 4 bar, initial liquid temperature of 20℃, nominal temperature of 25℃, and volumetric flow rate at zero differential pressure of 400m³ / h. 3 / h, nominal efficiency is 0.4.

[0290] ⑧ Building parameters: Room temperature is 26℃, and the load data is calculated manually.

[0291] (3) Model calibration and verification

[0292] First, August 22, 2024, was selected as the verification period, and the operating parameters of each component of the air conditioning system during this period were obtained through actual testing. Second, based on the obtained information, the operating parameters of the Modelica dynamic simulation model of the air conditioning system were set, and the air conditioning system model was simulated during this period. Finally, the simulated condenser inlet water temperature was compared with the inlet water temperature monitored by the actual system. The average deviation between the two during this period was calculated to be 10.6%, which is less than 15%. Therefore, it can be proved that the model can accurately reflect the operation of the actual air conditioning system.

[0293] (4) Fault Injection and Simulation

[0294] Obtaining fault data in actual engineering projects faces significant challenges:

[0295] (1) Active injection faults, such as blockage of pipelines causing abnormal reduction in chilled water flow or throttle valve stuck in an abnormal position, are often destructive and may cause permanent damage to equipment or even system shutdown, resulting in high maintenance costs and even safety hazards. They are usually not feasible.

[0296] (2) Passive collection of natural faults has a long cycle and limited sensor deployment, resulting in problems such as insufficient monitoring parameters, uncontrollable fault degree, inability to identify some faults in the early stage, and poor data quality.

[0297] Against this backdrop, the dynamic simulation model of a complex building air conditioning system built using the Modelica language in this embodiment can realistically reflect the complex dynamic behavior of the air conditioning system and output highly reliable simulation results. By injecting faults into this model, fault simulations of different severity levels and types can be achieved, generating high-quality, complete sets of air conditioning system operation data at different levels (component-equipment-system) under specific fault severity conditions. The specific fault injection and simulation steps are as follows:

[0298] (1) Determine the fault scenario: Determine the target fault type (single fault / multiple faults) and fault severity. This embodiment selects three fault types for simulation: condenser scaling, expansion valve sticking, and fan failure. Each type of fault is divided into three fault levels, corresponding to different degrees of severity. Under normal operating conditions, the model parameter corresponding to the fault is X. When controlling the fault severity to levels one, two, and three, the parameter is modified to 80%X, 60%X, and 40%X, respectively. Some fault scenario settings are as follows:

[0299] Table 2 shows some fault scenario settings in the embodiments.

[0300]

[0301]

[0302] (2) Identify faulty components and parameters: Based on the fault type, identify the faulty components and their corresponding model parameters. In this embodiment, the condenser scaling fault occurs in the condenser, and the fault can be injected by reducing the heat transfer system of the condenser to simulate the scenario where the heat transfer resistance increases due to the increase of scale or fouling layer; the throttle valve jamming fault occurs in the throttle valve, and the fault can be injected by fixing the effective flow area of ​​the throttle valve to simulate the scenario where the valve cannot be adjusted normally due to mechanical jamming, signal failure, etc.; the fan fault occurs in the fan, and the fault can be injected by reducing the mass flow rate of the fan to simulate insufficient air volume caused by motor failure, belt slippage, blade damage, etc.

[0303] (3) Fault injection and simulation: Based on steps (1) and (2), modify the parameter values ​​of the corresponding components in the Modelica dynamic simulation model of the air conditioning system to complete the injection of specific fault types and fault degrees, and perform fault simulation of the air conditioning system. For example, in the dual-fault concurrent scenario 2 in the embodiment, a first-level condenser scaling fault and a second-level throttle valve jamming fault occur simultaneously. Fault injection can be achieved by modifying the condenser heat transfer coefficient to 80% of the normal condition and the effective flow area of ​​the throttle valve to 80% of the normal condition, and then performing fault simulation.

[0304] (4) Output system fault data: After completing the fault simulation, output high-quality "component-equipment-system" fault operation data at different levels of the air conditioning system under specific fault scenarios, including parameters that are difficult to monitor in actual air conditioning systems, such as compressor power, evaporation pressure, condensation pressure, condenser heat dissipation power, and evaporator cooling power.

[0305] The above are only some preferred embodiments of this application, but this application is not limited thereto, and many improvements and modifications can be made. Any improvements and modifications made based on the basic principles of this application should be considered to fall within the protection scope of this application.

Claims

1. A method for simulating a failure of a complex air conditioning system based on a Modelica language, characterized by, The steps include the following: S1, Air Conditioning System Model Establishment: Based on the Modelica language, establish multiple general component models of the air conditioning system, add corresponding connectors between the multiple general component models to establish multiple key equipment models, and connect the multiple key equipment models to form the Modelica dynamic simulation model of the air conditioning system. S2, Parameter Acquisition and Setting: Acquire parameter information of each device in the actual air conditioning system, and set the Modelica dynamic simulation model of the air conditioning system based on the parameter information; S3, Model Calibration and Verification: Configure the Modelica dynamic simulation model of the air conditioning system according to the parameter information of each device, and perform model calibration and verification by comparing the simulation output data with the actual monitoring data until the deviation between the two is less than or equal to the preset threshold. S4, Fault Injection and Simulation: Based on fault scenario identification of faulty equipment and parameter information, modify the parameter values ​​of corresponding components in the Modelica dynamic simulation model of the air conditioning system to inject faults, and output a complete set of air conditioning system fault simulation output data under specific fault scenarios at the system level.

2. The method of claim 1, wherein, In S1: The various general component models specifically include heat exchanger models, compressor models, expansion valve models, fan models, and surface cooler models; The specific models of the key equipment include chiller models, air handling unit models, fan coil unit models, cooling tower models, chilled pump models, and cooling pump models.

3. The method of claim 1, wherein, In step S1, the multiple key equipment models are connected according to the physical topology and thermodynamic cycle relationship of the actual air conditioning system to form a complete Modelica dynamic simulation model of the air conditioning system.

4. The method of claim 1, wherein, In step S1, based on the working principle of the chiller unit and the connection relationship of each component, the chiller unit model is established based on the heat exchanger model, compressor model and throttle valve model.

5. The method of claim 1, wherein, In S1, the air handling unit model, fan coil unit model, and cooling tower model are all based on the general structure of "fan model + surface cooler model".

6. The method of claim 1, wherein, In step S2, the parameter information of each device includes: Heat exchanger: heat transfer coefficient, initial liquid volume fraction, initial liquid temperature, initial wall temperature; Compressor: speed, displacement, volumetric efficiency, isentropic efficiency, effective isentropic efficiency; Throttling valve: effective flow area; Chiller units: cooling water flow rate, chilled water flow rate; Fan: mass flow rate, fan efficiency, drive efficiency; Surface cooler: heat transfer coefficient, initial liquid temperature, initial wall temperature, initial gas pressure drop; Refrigeration pump / cooling pump: speed, pressure increment at zero volume flow rate, initial liquid temperature, nominal temperature, volume flow rate at zero differential pressure, nominal efficiency; Building parameters: room temperature, load data.

7. The method of claim 1, wherein, S3 specifically includes: S31: Select a verification time period and obtain the values ​​of the parameter information of each device within the verification time period; S32: Set the parameters of the Modelica dynamic simulation model of the air conditioning system according to the parameter values, and simulate the simulation output data of each device during the verification period. S33: Compare the simulation output data with the actual monitoring data. If the deviation is less than or equal to the preset threshold, it means that the model can reflect the actual operation of the air conditioning system well and no adjustment is needed. If the deviation is greater than the preset threshold, the key parameters need to be readjusted until the deviation is less than the preset threshold.

8. The method according to claim 1, characterized in that, S4 specifically includes: S41: Determine the fault scenario: Determine the fault type and fault severity, wherein the target fault type includes single fault and multiple fault; S42: Identify faulty equipment and parameter information: Based on the fault type, identify the faulty equipment and its corresponding parameter information; S43: Fault Injection and Simulation: Based on the faulty equipment and its corresponding parameter information, set the parameter values ​​of the corresponding components in the Modelica dynamic simulation model of the air conditioning system, complete the injection of specific fault types and fault degrees, and simulate the faults of the air conditioning system. S44: Output system fault data: Output system-level, complete set of air conditioning system fault simulation output data under specific fault scenarios.

9. The method according to claim 8, characterized in that, In step S42, based on the fault type, the faulty device and its corresponding parameter information are identified, specifically including multiple triplets consisting of (fault type, faulty component, corresponding model parameter): (Condenser scaling, condenser, condenser heat transfer coefficient) (Evaporator scaling, evaporator, evaporator heat transfer coefficient) (Compressor malfunction, compressor speed) (Throttle valve stuck, throttle valve, effective flow area) (Fan malfunction, fan, fan mass flow rate) (Scale or blockage in the surface cooler, surface cooler, surface cooler heat transfer coefficient) (Abnormal cooling water flow rate, chiller unit, cooling water mass flow rate) (Abnormal chilled water flow rate, chiller unit, chilled water mass flow rate).

10. A fault simulation system for a complex air conditioning system based on the Modelica language, said system operating according to the method according to any one of claims 1 to 9, wherein, The system includes: An air conditioning system model building unit is used to build multiple general component models of an air conditioning system based on the Modelica language, add corresponding connectors between the multiple general component models to build multiple key equipment models, and connect the multiple key equipment models to form a Modelica dynamic simulation model of the air conditioning system. The parameter acquisition and setting unit is used to acquire parameter information of each device in the actual air conditioning system and set the Modelica dynamic simulation model of the air conditioning system based on the parameter information. The model calibration and verification unit is used to configure the Modelica dynamic simulation model of the air conditioning system according to the parameter information of each device, and to perform model calibration and verification by comparing the simulation output data with the actual monitoring data until the deviation between the two is less than or equal to a preset threshold. The fault injection and simulation unit is used to identify faulty equipment and parameter information based on fault scenarios, modify the parameter values ​​of corresponding components in the Modelica dynamic simulation model of the air conditioning system, inject faults, and output a complete set of air conditioning system fault simulation output data under specific fault scenarios at the system level.

Images