Air-conditioning machine room performance dynamic optimization control method and system based on ESN and SSA

Abstract

The application provides an air-conditioning machine room performance dynamic optimization control method and system based on ESN and SSA, and the method comprises the following steps: collecting real-time operation data of an air-conditioning machine room; inputting the real-time operation data of each device into a pre-constructed ESN model of each device; dynamically generating operation stability of each device according to the real-time operation data; combining the ESN energy consumption model and the operation stability of each device to construct a multi-objective optimization function, searching for an optimization control strategy of the device by using an SSA algorithm; executing the optimization control strategy, collecting feedback data of the device, rolling updating an input sequence of the pre-constructed ESN energy consumption model and performing model updating iteration, and forming a control strategy for a next control period. The application constructs a multi-objective optimization framework which simultaneously satisfies system total power minimization and device operation stability maximization, and can keep the air-conditioning machine room energy efficiency and operation stability optimal under the condition that the air-conditioning machine room load and device state dynamically change.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of air-conditioning machine room performance optimization, and particularly relates to an air-conditioning machine room performance dynamic optimization control method and system based on ESN and SSA. BACKGROUND

[0002] As the core facility of large buildings, industrial parks and data centers, the energy efficiency and operation stability of air-conditioning machine rooms directly affect the energy consumption and maintenance cost of the whole building. In actual operation, air-conditioning machine rooms need to coordinate multiple chillers, water pumps, cooling water pumps and cooling towers to meet the dynamic cooling load demand. However, the system operation process is often accompanied by significant thermal inertia effect and device response delay, making it difficult for traditional control methods relying on rule base or linear regulation to meet the dynamic changing load demand with the lowest energy consumption. In addition, after running for a period of time, devices such as chillers, water pumps and cooling towers may overheat, mechanically fatigue and performance decline. Frequent, severe fluctuations in the working current of the device or high load can reduce the operation stability of the device. The existing control method only considers the adjustment of the device according to the change of the load, lacks dynamic evaluation of the running health status of the device, ignores the running state and performance change of the air-conditioning machine room device, and may cause device overload, frequent start-stop or energy efficiency reduction, or even device failure. Moreover, the control execution result often fails to effectively feedback into the next cycle modeling process, lacking a sustainable and adaptive rolling closed-loop updating mechanism, which limits the robustness and long-term optimization ability of the data-driven control system in actual complex environment.

[0003] Therefore, the current intelligent control system of air-conditioning machine rooms needs to meet the dynamic load demand with the lowest energy consumption, sense the running state and performance change of the device and make corresponding control decisions, and have a periodical rolling execution and feedback ability. SUMMARY

[0004] The purpose of the present application is to provide an air-conditioning machine room performance dynamic optimization control method and system based on ESN and SSA to solve the above problems.

[0005] In order to achieve the above purpose, in the first aspect of the present application, an air-conditioning machine room performance dynamic optimization control method based on ESN and SSA is provided, which comprises the following steps:

[0006] Collecting real-time operation data of the air-conditioning machine room and pre-processing the original data; wherein the real-time operation data includes the working temperature and working current of the chiller compressor, water pump and cooling tower motor;

[0007] Inputting the pre-processed real-time operation data into the pre-constructed ESN model of each device to predict the power of the chiller compressor, water pump and cooling tower motor;

[0008] According to the working temperature and working current of the water chiller compressor, water pump and cooling tower motor, the running stability of each device is dynamically generated;

[0009] Combined with the power and running stability of the water chiller compressor, water pump and cooling tower motor, a multi-objective optimization function is constructed, and the SSA algorithm is used to search for the optimal running parameters and start-up combination of the device as the optimized control strategy of the device;

[0010] The optimized control strategy is executed, real-time device feedback data is collected, the pre-constructed ESN energy consumption model input sequence is updated and model update iteration is performed for the control strategy formation in the next control period.

[0011] Further, the real-time running data of the air conditioning machine room is collected, and the pre-processing step of the original data includes:

[0012] The power, cold water flow, cold water supply and return water temperature, cooling water flow and cooling water inlet temperature of each water chiller are collected; the power, cooling water flow, water inlet and outlet temperature, air dry bulb temperature and wet bulb temperature of the cooling tower are collected; the power and flow of the water pump are collected; the working temperature of the water chiller compressor, water pump and cooling tower motor is collected, and the working current value of each device is collected.

[0013] For the original data collected by the sensor, abnormal data is removed;

[0014] According to the cold water flow, cold water supply and return water temperature, the current cooling load is calculated, and according to the cooling water flow, cooling tower water inlet and outlet temperature, the current cooling load of the cooling tower is calculated;

[0015] The real-time state data and historical state sequence of each device are arranged based on a unified time window format to generate a multi-dimensional time sequence.

[0016] Further, the pre-constructed ESN model of each device of the air conditioning machine room performance dynamic optimization control based on ESN and SSA consists of an input layer, a fixed randomly connected Reservoir layer and a trained output layer;

[0017] The input parameters of the input layer are: the cooling load, cold water supply temperature, cold water flow, cooling water flow and cooling water inlet temperature required by the water chiller energy consumption model; the cooling load, air dry bulb temperature, air wet bulb temperature, cooling water flow and outlet temperature required by the cooling tower energy consumption model; the water pump flow required by the water pump energy consumption model.

[0018] The Reservoir layer is used to map the input to a nonlinear, high-dimensional dynamic space and generate the current state vector of the Reservoir.

[0019] The output layer maps the current state vector of the Reservoir to the predicted value of the future target variable, namely the power of each device;

[0020] The pre-built ESN model only adjusts the output layer weight matrix during the model training phase, using recursive least squares to train the output layer weights, eliminating the need to train the Reservoir layer. Furthermore, the operational stability of the air conditioning room performance dynamic optimization control based on ESN and SSA is used to measure the stability of equipment operation. The formula for calculating the operational stability of the i-th device is as follows:

[0021] ;

[0022] in: The ratio of the current operating temperature to the maximum allowable operating temperature of the i-th device reflects the operating temperature level of the device. , which is the maximum value of the rate of change of current, reflects the intensity of current fluctuations over a short period of time; The ratio of the current average current to the maximum operating current of the i-th device reflects the current load of the device. These are the weighting coefficients.

[0023] Furthermore, the multi-objective optimization function for dynamic performance optimization control of the air conditioning room based on ESN and SSA, which aims to simultaneously minimize the total system power and maximize the equipment operating stability, is expressed as:

[0024] ;

[0025] ;

[0026] Where P: Total power of all devices; : No. The start / stop status of the chiller unit and its corresponding water pump, S i =1 or 0; : No. The start / stop status of the cooling tower, Z i =1 or 0; W i P i C i T i : These represent the power of the i-th chiller unit, chilled water pump, cooling water pump, and cooling tower, respectively; B: The sum of the operational stability of all devices; i D i F i, G i : running stability of the i-th chiller, chiller pump, cooling water pump, and cooling tower, respectively; N: total number of any one of the four types of equipment, 1≤i≤N.

[0027] Further, the constraint conditions of the multi-objective optimization function are: (1) the sum of the cooling loads borne by each chiller is equal to the system load; (2) at any time, at least one cooling water pump, one chiller, and one cooling tower are started, and the number of started cooling towers is not less than the number of started chillers; (3) each device cannot be turned off within half an hour after being started, and each device cannot be started within half an hour after being turned off.

[0028] Further, the multi-objective optimization function of the air conditioning machine room performance dynamic optimization control based on ESN and SSA is searched and solved by using a sparrow search algorithm, and the search space of the sparrow search algorithm is a variable vector with a dimension of In the initialization stage, the discoverer group is set to guide the target value, and the follower group is set to perform local search; in each iteration, the search strategy is updated according to the current optimal target value until the convergence threshold is reached or the upper limit of the number of iterations is reached.

[0029] Further, the air conditioning machine room performance dynamic optimization control based on ESN and SSA executes the optimization control strategy, collects real-time device feedback data, and updates the pre-constructed ESN energy consumption model input sequence and performs model update iteration, which are used for the control strategy formation in the next control period.

[0030] The optimization control strategy is executed through the existing PLC system of the air conditioning machine room;

[0031] The PLC system startup state confirmation mechanism: after each device issues a control instruction, the system waits for 15-30 seconds to check the return state variable; if the device responds normally, it is recorded as a successful control; if the state does not change or the parameter is abnormal, it is marked in the abnormal device list;

[0032] Collect the state data of each device after executing the control, form a state vector with the state data, splice it with the historical state window, and construct an input sequence of the pre-constructed ESN model for model update iteration.

[0033] Further, when the device executes the control based on ESN and SSA, each instruction is packaged according to the following structure:

[0034] Start / stop command: write to the address bit corresponding to the device;

[0035] Frequency command: write to the frequency setting address of the frequency converter.

[0036] In a second aspect of the present application, an ESN and SSA based air conditioning machine room performance dynamic optimization control system is provided, the system comprising:

[0037] A data acquisition and processing module: acquiring real-time operation data of the air conditioning machine room, and pre-processing the original data; wherein the real-time operation data includes working temperature and working current of the chiller compressor, water pump and cooling tower motor;

[0038] An equipment power prediction module: for inputting the pre-processed real-time operation data into the pre-constructed ESN model of each device, and predicting the power of the chiller compressor, water pump and cooling tower motor;

[0039] An equipment operation stability evaluation module: for dynamically generating the operation stability of each device according to the working temperature and working current of the chiller compressor, water pump and cooling tower motor;

[0040] A control strategy optimization module: combining the power and operation stability of the chiller compressor, water pump and cooling tower motor, constructing a multi-objective optimization function, and searching for the optimal operation parameters and start-up combination of the equipment by using the SSA algorithm as the optimized control strategy of the equipment;

[0041] A rolling update module: executing the optimized control strategy, collecting real-time equipment feedback data, rolling updating the pre-constructed ESN energy consumption model input sequence and performing model update iteration, for forming the control strategy of the next control period.

[0042] The present application has at least the following beneficial technical effects:

[0043] The present application provides a performance dynamic optimization control method and system in view of the dynamic change characteristics of air conditioning machine room load and equipment state, which can not only meet the dynamic load needs with the lowest energy consumption, but also can perceive the equipment operation state and performance change and maximize the equipment operation stability, so that the machine room energy efficiency and equipment operation stability are both kept optimal under the dynamic change of machine room load and equipment state.

[0044] The present application proposes the equipment operation stability of air conditioning machine room and its calculation method, which provides a quantitative basis for ensuring stable operation of equipment in the air conditioning machine room control process, and overcomes the problems of equipment overload, frequent start-stop, reduced energy efficiency or faults caused by ignoring the change of equipment operation state in the existing control method.

[0045] This invention constructs a multi-objective optimization function that simultaneously minimizes the total system power and maximizes the equipment's operational stability. It then utilizes a sparrow search algorithm to search for the optimal control strategy and equipment start-up combination in the discrete start-up and continuous frequency control space. This achieves the organic unity of coordinated optimization control and stable equipment operation among multiple chiller units and corresponding water pumps and cooling towers in large and medium-sized air conditioning rooms.

[0046] The control system provided by this invention generates control commands that are sent to the field PLC system through a standard communication interface. After executing the control commands, feedback data is collected in real time, and the system input status window and the device ESN energy consumption model are updated to drive the next control cycle to run in a rolling manner, forming a stable and sustainable "decision-execution-feedback" closed loop. Attached Figure Description

[0047] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0048] Figure 1 This is a flowchart of the dynamic performance optimization control method for air conditioning computer room based on ESN and SSA of the present invention;

[0049] Figure 2 This is a framework diagram of the dynamic optimization control system for air conditioning room performance based on ESN and SSA of the present invention. Detailed Implementation

[0050] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0051] The ESN model is an energy consumption model for echo state networks.

[0052] The SSA algorithm is a sparrow search algorithm.

[0053] like Figure 1 As shown in the embodiment of the present invention, a dynamic optimization control method for the performance of air conditioning room equipment based on ESN and SSA is provided. The method includes:

[0054] S1, data acquisition and data processing. The power, cold water flow, cold water supply and return water temperature, cooling water flow, cooling water inlet temperature of each cold water main are collected; the power, cooling water flow, water inlet and outlet temperature, inlet air dry bulb temperature and wet bulb temperature of the cooling tower are collected; the power and flow of the water pump are collected; the working temperature of the compressor, water pump and cooling tower motor of the cold water unit is collected, and the working current value of each device is collected. The current cooling load is calculated according to the cold water flow, cold water supply and return water temperature, and the current cooling load is calculated according to the cooling water flow, cooling tower water inlet and outlet temperature. For the data collected by the sensor, the abnormal data is removed, and then the average value of the original data is calculated according to a certain sampling period. All input data are collected by the existing sensors in the machine room, without additional equipment modification. The sampling period of data can be 1 minute, and the historical window length can be 30, covering 30 minutes of state evolution.

[0055] For the data collected by the sensor, the median value filtering method is used to remove abnormal data generated by external interference to the sensor.

[0056] Further, the collected data is arranged into a unified time window format to form a multi-dimensional time series input:

[0057]

[0058] wherein, represents the device running data at time point ; n is the size of the time window.

[0059] S2, input the collected data of each device into the pre-constructed ESN energy consumption model of each device for predicting the power of each device. The echo state network (ESN) has high training efficiency and is suitable for nonlinear system modeling. The ESN energy consumption model of the device is composed of an input layer, a fixed random connected Reservoir layer and a trained output layer. The input parameters of the input layer are: the cold water unit energy consumption model needs to input the cooling load, cold water supply temperature, cold water flow, cooling water flow, cooling water inlet temperature; the cooling tower energy consumption model needs to input the cooling load, inlet air dry bulb temperature, inlet air wet bulb temperature, cooling water flow, outlet water temperature; the water pump energy consumption model needs to input the water pump flow.

[0060] The Reservoir layer is used to map the input to a nonlinear, high-dimensional dynamic space, and its state recursive formula is as follows:

[0061]

[0062] wherein, is the current state vector of the Reservoir, is the input weight matrix, ​​is the state feedback matrix, is the bias term, is a nonlinear activation function (e.g., tanh). This structure captures the inertia characteristics in time series through internal dynamics.

[0063] Subsequently, the output layer maps the Reservoir states to the predicted values of the future target variable:

[0064] ;

[0065] where, is the predicted device power vector, is the output layer weight matrix. During the model training phase, the output layer weights are trained using recursive least squares The neuron connections and weights of the Reservoir layer are fixed and do not require training.

[0066] S3, according to the working temperature and working current of the compressor, water pump and cooling tower motor of the water chiller unit, dynamically generates the running stability of each device.

[0067] Specifically, the operation optimization control of the air conditioning machine room not only requires the minimization of the total power of each device, but also requires good running stability of the machine room equipment. After running for a period of time, devices such as water chillers, water pumps and cooling towers will overheat, mechanical fatigue and performance decline. Frequent, severe fluctuations in device working current or high load will reduce the running stability of the device. If the changes in device status and performance are ignored, it may lead to device overload, frequent start-stop or reduced energy efficiency, and even failure.

[0068] Further, in order to measure the stability of the device operation, we define the device operation stability The higher the value, the more stable the device operation and the more suitable for regulation and control; the lower the value, the more unstable the device operation and the less suitable for regulation and control. The calculation formula of the running stability of the i-th device is as follows:

[0069] ;

[0070] where: is the ratio of the current working temperature of the i-th device to the maximum allowable working temperature, reflecting the working temperature level of the device; is the maximum value of the current change rate, reflecting the intensity of current fluctuation in a short period of time; : the ratio of the current average current of the i-th device to the maximum working current, reflecting the current load of the device; is a weighting coefficient, which can be set according to the experience of device type, for example , , It should be noted that the formula adopts an inverse structure, ensuring that the stability of the equipment operation decreases when each index increases. When optimizing the control, the operation of the equipment with excessively high working temperature, severe current fluctuation, or excessively high load is inhibited as much as possible to improve the stability of the operation of the air-conditioning machine room equipment.

[0071] S4, in combination with the ESN energy consumption model and the operation stability of the equipment, a multi-objective optimization function is constructed, and a sparrow search algorithm is used to search for optimal operation parameters and start combinations of the equipment as the optimization control strategy of the equipment.

[0072] Specifically, after the power prediction of each device (step S2) and the operation stability evaluation of each device (step S3) are completed, the control system needs to generate a device optimization control strategy of the whole system under the constraint condition, reasonably set the start-stop state and operation frequency of each device, and enable the system to realize optimal performance operation under dynamic working conditions. Unlike the experience rule or PID adjustment mode in the traditional air-conditioning refrigeration control system, the present application scheme proposes a multi-objective optimization method based on a sparrow search algorithm (Sparrow Search Algorithm, SSA), which integrates the maximum energy efficiency of the machine room equipment and the operation stability of the machine room equipment into a unified optimization framework.

[0073] Further, the optimization target is to generate a control instruction combination , wherein represents the start-stop state of the i-th device, 1 represents start, and 0 represents stop; represents the operation frequency thereof. The optimization control needs to simultaneously satisfy the minimum system total power and the maximum equipment operation stability. Based on the above consideration, the following multi-objective optimization function is constructed:

[0074] ;

[0075] ;

[0076] , wherein the constraint conditions are as follows: (1) the sum of the cold loads borne by each chiller unit is equal to the system load; (2) at any time, at least one cooling water pump, one chiller unit, and one cooling tower need to be started, and the number of started cooling towers is not less than the number of started chiller units; (3) each device cannot be stopped within half an hour after being started, and each device cannot be started within half an hour after being stopped. Wherein, P: total power of all devices; : start-stop state of the i-th chiller unit and the water pump corresponding thereto; : start-stop state of the i-th cooling tower; W : start-stop state of the i-th chiller unit and the water pump corresponding thereto; : start-stop state of the i-th cooling tower; W i , P i , C i , T​i : the power of the i-th chiller, chiller pump, cooling water pump and cooling tower, respectively; : the sum of the operating stability of all devices;B i , D i , F i , G i : the operating stability of the i-th chiller, chiller pump, cooling water pump and cooling tower, respectively; N: the total number of any one of the four types of devices, i.e. chiller, chiller pump, cooling water pump and cooling tower, 1≤i≤N.

[0077] It should be noted that by searching for the above multi-objective mixed integer programming problem, the chiller pump and cooling water pump flow setting value, cooling tower outlet water temperature setting value and device on-off combination that simultaneously minimize the total power of the system and maximize the operating stability of the device can be obtained.

[0078] The search process of the control strategy is realized by the sparrow search algorithm (SSA). The search space of SSA is a variable vector of dimension , i.e. , where the first dimensions are discrete variables and the last dimensions are continuous variables. In the initialization stage of the algorithm, the discoverer population is guided by the target value, and the follower population is searched locally. In each iteration, the search strategy is updated according to the current optimal target value until the convergence threshold is reached or the maximum number of iterations is reached.

[0079] The advantage of SSA is its simple structure and controllable parameters, which is particularly suitable for multi-objective optimization problems such as refrigeration systems, which have a fixed number of control variables, mixed continuous and discrete variables.

[0080] After searching and solving, the optimal on-off combination and the optimal chiller flow, cooling water flow and cooling tower outlet water temperature setting value are obtained. The optimal setting value is converted into the running frequency setting value of the water pump and cooling tower fan, and finally output as the device control instruction combination: , i.e. the start-stop state and running frequency setting value of each device.

[0081] S5, execute the optimization control strategy, collect device feedback data, rollingly update the pre-constructed ESN energy consumption model input sequence and perform model update iteration for the control strategy formation of the next control period. Specifically, the purpose of this step is to update the optimal device control strategy generated in the previous step The control execution system is actually issued to the air conditioning room, and the device feedback state is taken as the input of the next round of prediction model, so as to construct the "perception-evaluation-decision-execution-re-perception" closed-loop control process of the system. In order to ensure the implementability of the strategy, the control instruction standardization interface, the device response confirmation mechanism and the rolling input update mechanism are designed in this step, so as to ensure that the system can complete the strategy application and data backwrite in each control period.

[0082] Get: : indicates whether the device is enabled (1 for start, 0 for stop); : the running frequency setting value (Hz) of the device ; the combination is represented as control combination .

[0083] Further, the execution of the control strategy is realized through the existing PLC system of the air conditioning room. The PLC has integrated device control channels and data acquisition channels, and the system is connected with devices such as water chillers, pumps and cooling towers through industrial bus protocols (such as ModbusTCP or BACnet).

[0084] When executing the control, each instruction is packaged according to the following structure:

[0085] Start-stop command: write the corresponding address bit of the device, such as Device[i].RunCmd=s_i^*;

[0086] Frequency command: write the frequency setting address of the frequency converter, such as Device[i].FreqSet=f_i^*.

[0087] Taking an example, if the third water chiller , Hz, the system will write the following two instructions to its PLC channel:

[0088] Chiller[3].Start=TRUE

[0089] Chiller[3].VFD_Frequency=45.0

[0090] The system then starts the state confirmation mechanism: after issuing the control instruction to each device, the system waits for 15-30 seconds, checks the returned state variables, such as the running indication bit RunStatus=1 and the current acquisition value I>0.2A. If the device response is normal, it is recorded as "control success"; if the state does not change or the parameters are abnormal (such as under-voltage and tripping), it is marked in the abnormal device list and fed back to the state update module.

[0091] Subsequently, the state data of each device after the control is completed is collected as the latest frame of the input sequence of the next round of ESN model, spliced with the historical state window to form a new time sequence input. Through this way, a rolling update mechanism is constructed, after the end of each round of control, the system can automatically collect new state feedback and drive model update, for the control strategy formation of the next control period.

[0092] As shown in Figure 2 The embodiment of the application also provides an air-conditioning machine room equipment performance dynamic optimization control system based on ESN and SSA, which comprises:

[0093] A data acquisition and processing module 201 is configured to collect real-time operation data of the air-conditioning machine room and pre-process the original data, wherein the real-time operation data comprises working temperature and working current of a chiller compressor, a water pump and a cooling tower motor.

[0094] An equipment power prediction module 202 is configured to input the pre-processed real-time operation data into pre-constructed ESN models of each device to predict power of the chiller compressor, the water pump and the cooling tower motor.

[0095] An equipment operation stability evaluation module 203 is configured to dynamically generate operation stability of each device according to the working temperature and the working current of the chiller compressor, the water pump and the cooling tower motor.

[0096] A control strategy optimization module 204 is configured to combine the power and the operation stability of the chiller compressor, the water pump and the cooling tower motor, construct a multi-objective optimization function, and search for optimal operation parameters and start-up combinations of the equipment by using an SSA algorithm to serve as an optimized control strategy of the equipment.

[0097] A rolling update module 205 is configured to execute the optimized control strategy, collect real-time equipment feedback data, rollingly update the pre-constructed ESN energy consumption model input sequence and perform model update iteration, for control strategy formation of the next control period.

[0098] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the system, device and unit described above can refer to the corresponding process in the foregoing method embodiments, which will not be described here.

[0099] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented in other manners. For example, the division of the above-described device embodiments is merely a logical function division, and there can be another division manner for the actual implementation, for example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections between different units, or the among different units, can be indirect couplings or communication connections through some interfaces, devices or units, and can be in electrical, mechanical or other forms.

[0100] The functions described above can be implemented in software, firmware, hardware, or any combination thereof. Moreover, the functions will be implemented in software as functions of an application program running on a computer, in one embodiment. Furthermore, if implemented in software, the functions can be stored in or transmitted over as one or more instructions or code on a computer-readable medium. Computer-readable media include both computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. A storage media can be any available media that can be accessed by a computer. By way of example, and not limitation, such computer-readable media can comprise RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer. Also, any connection is properly termed a computer-readable medium. For example, if the software is transmitted from a website, server, or other remote source using a coaxial cable, fiber optic cable, or twisted pair, then the coaxial cable, fiber optic cable, or twisted pair are included in the definition of medium. Disk and disc, as used herein, include compact disc (CD), laser disc, optical disc, digital versatile disc (DVD), and Blu-Ray® disc where disks usually reproduce data magnetically, while discs reproduce data optically with lasers. Combinations of the above should also be included within the scope of computer-readable media.

[0101] In light of the above, it should be appreciated that modifications and variations to the present application can be effected without departing from the scope of the application. Accordingly, the scope of the application should be judged in relation to the claims and equivalents thereof.

Claims

1. A dynamic optimization control method for air conditioning room performance based on ESN and SSA, characterized in that, The method includes the following steps: Collect real-time operating data of the air conditioning room and preprocess the raw data; wherein, the real-time operating data includes the operating temperature and operating current of the chiller compressor, water pump and cooling tower motor; The preprocessed real-time operating data is input into the pre-built ESN model of each device to predict the power of the chiller compressor, water pump and cooling tower motor; Based on the operating temperature and current of the chiller unit's compressor, water pump, and cooling tower motor, the operational stability of each device is dynamically generated. Combining the power and operational stability of the chiller unit's compressor, water pump, and cooling tower motors, a multi-objective optimization function is constructed. The SSA algorithm is then used to search for the optimal operating parameters and start-up combinations of the equipment, serving as the optimal control strategy for the equipment. The optimized control strategy is executed, and real-time device feedback data is collected. The pre-built ESN energy consumption model input sequence is updated and the model is updated iteratively for the formation of the control strategy in the next control cycle.

2. The method for dynamic optimization control of air conditioning room performance based on ESN and SSA according to claim 1, characterized in that, The steps for collecting real-time operating data from the air conditioning room and preprocessing the raw data include: Collect the power, chilled water flow rate, chilled water supply and return water temperature, cooling water flow rate, and cooling water inlet temperature of each chiller unit; collect the power, cooling water flow rate, inlet and outlet water temperature, inlet dry bulb temperature, and wet bulb temperature of the cooling tower; collect the power and flow rate of the water pump; collect the operating temperature of the chiller unit compressor, water pump, and cooling tower motor; and collect the operating current value of each device. For the raw data collected by the sensors, abnormal data is removed; The current cooling load is calculated based on the cold water flow rate, cold water supply and return water temperatures, and the current cooling load of the cooling tower is calculated based on the cooling water flow rate, cooling tower inlet and outlet water temperatures. The real-time status data and historical status sequences of each device are organized based on a unified time window format to generate a multi-dimensional time series.

3. The method for dynamic optimization control of air conditioning room performance based on ESN and SSA according to claim 1, characterized in that, The pre-built ESN model for each device consists of an input layer, a fixed randomly connected Reservoir layer, and a trained output layer. The input parameters for the input layer are as follows: for the chiller unit energy consumption model, the input parameters are: cooling load, chilled water supply temperature, chilled water flow rate, cooling water flow rate, and cooling water inlet temperature; for the cooling tower energy consumption model, the input parameters are: cooling load, inlet dry bulb temperature, inlet wet bulb temperature, cooling water flow rate, and outlet temperature; and for the water pump energy consumption model, the input parameter is: water pump flow rate. The Reservoir layer is used to map the input to a nonlinear, high-dimensional dynamic space and generate the current state vector of the Reservoir. The output layer maps the current state vector of the Reservoir to the predicted value of the future target variable, namely the power of each device; During the model training phase of the pre-built ESN model, only the output layer weight matrix is ​​adjusted, and the output layer weights are trained using recursive least squares method, without the need to train the Reservoir layer.

4. The method for dynamic optimization control of air conditioning room performance based on ESN and SSA according to claim 1, characterized in that, The operational stability is used to measure the stability of equipment operation. The formula for calculating the operational stability of the i-th device is as follows: ； in: The ratio of the current operating temperature to the maximum allowable operating temperature of the i-th device reflects the operating temperature level of the device. , which is the maximum value of the rate of change of current, reflects the intensity of current fluctuations over a short period of time; The ratio of the current average current to the maximum operating current of the i-th device reflects the current load of the device. These are the weighting coefficients.

5. The method for dynamic optimization control of air conditioning room performance based on ESN and SSA according to claim 1, characterized in that, The multi-objective optimization function aims to simultaneously minimize the total system power and maximize the equipment operating stability, and is expressed as follows: ； ； Where P: Total power of all devices; : No. The start / stop status of the chiller unit and its corresponding water pump, S i =1 or 0; : No. The start / stop status of the cooling tower, Z i =1 or 0; W i P i C i T i : These represent the power of the i-th chiller unit, chilled water pump, cooling water pump, and cooling tower, respectively; B: The sum of the operational stability of all devices; i D i F i G i : represents the operational stability of the i-th chiller unit, chilled water pump, cooling water pump, and cooling tower, respectively; N: the total number of any one of the four types of equipment, 1≤i≤N.

6. The method for dynamic optimization control of air conditioning room performance based on ESN and SSA according to claim 5, characterized in that, The constraints of the multi-objective optimization function are: (1) the sum of the cooling loads borne by each chiller unit is equal to the system load; (2) at any time, at least one cooling water pump, one chiller unit and one cooling tower must be turned on, and the number of cooling towers turned on shall not be less than the number of chiller units turned on; (3) each equipment cannot be turned off within half an hour after it is turned on, and each equipment cannot be turned on within half an hour after it is turned off.

7. The method for dynamic optimization control of air conditioning room performance based on ESN and SSA according to claim 1, characterized in that, The multi-objective optimization function is solved using the sparrow search algorithm, and the search space of the sparrow search algorithm is dimensional. The variable vector; in the initialization phase, the discoverer group is set to guide the target value, and the follower group performs local search; in each iteration, the search strategy is updated according to the current optimal target value until the convergence threshold or the upper limit of the number of iterations is reached.

8. The method for dynamic optimization control of air conditioning room performance based on ESN and SSA according to claim 1, characterized in that, The steps for executing the optimized control strategy, collecting real-time device feedback data, continuously updating the pre-built ESN energy consumption model input sequence, and performing model update iterations to form the control strategy for the next control cycle include: The optimized control strategy is executed through the existing PLC system in the air conditioning room. The PLC system startup status confirmation mechanism is as follows: after each device issues a control command, the system waits for 15 to 30 seconds and checks the returned status variables; if the device responds normally, it is recorded as a successful control; if the status does not change or the parameters are abnormal, it is marked in the list of abnormal devices. Collect the status data of each device after the control is completed, form a status vector from the status data, and concatenate it with the historical status window to construct the input sequence of the pre-built ESN model for model update iteration.

9. The method for dynamic optimization control of air conditioning room performance based on ESN and SSA according to claim 8, characterized in that, When the device performs control, each instruction is encapsulated in the following structure: Start / Stop command: Write to the corresponding address bits of the device; Frequency command: Write the frequency setting address to the inverter.

10. A dynamic optimization control system for air conditioning room performance based on ESN and SSA, characterized in that, The system includes: Data acquisition and processing module: collects real-time operating data of the air conditioning room and preprocesses the raw data; wherein, the real-time operating data includes the operating temperature and operating current of the chiller compressor, water pump and cooling tower motor; Equipment power prediction module: used to input the pre-processed real-time operating data into the pre-built ESN model of each device to predict the power of the chiller unit compressor, water pump and cooling tower motor; Equipment operation stability assessment module: used to dynamically generate the operation stability of each piece of equipment based on the operating temperature and operating current of the compressor, water pump and cooling tower motor of the chiller unit; Control strategy optimization module: Combining the power and operational stability of the chiller unit's compressor, water pump, and cooling tower motors, a multi-objective optimization function is constructed. The SSA algorithm is used to search for the optimal operating parameters and start-up combinations of the equipment, which serve as the optimized control strategy for the equipment. Rolling update module: Executes the optimized control strategy, collects real-time device feedback data, and continuously updates the pre-built ESN energy consumption model input sequence and performs model update iterations for the formation of the control strategy in the next control cycle.

Images