An Adaptive Optimization Control Method for Chiller Systems

Abstract

This invention discloses an adaptive optimization control method for a chiller system. The method includes: in a computer room system containing at least two chillers, establishing and dynamically updating a chiller performance monitoring relationship table based on real-time collected operating data; and determining a load allocation operation scheme for the chiller that meets the total cooling load demand and has the highest system performance coefficient based on the chiller dynamic performance monitoring relationship table. This method can incorporate the influence of operating parameters such as chilled water inlet temperature, chilled water flow rate, cooling water inlet temperature, and cooling water flow rate, and can adapt to situations where chiller performance parameters are unknown or deviate. It offers high accuracy and reliability; adaptive adjustments can be made when the number of chillers changes; it is adaptable to different computer rooms and has strong versatility.

Description

Technical Field

[0001] This invention relates to the field of optimized operation control technology for chiller rooms, specifically to an adaptive optimization control method for a chiller unit system. Background Technology

[0002] Air conditioning and refrigeration rooms have historically suffered from low energy efficiency, leaving significant room for energy conservation. Chillers are the core equipment in these rooms, consuming the most energy. Typically, a refrigeration system consists of multiple chiller units, each with varying performance characteristics. Furthermore, the performance of these chiller units changes under different operating conditions (including unit load rate, chilled water inlet temperature and flow rate, and cooling water inlet temperature and flow rate). Therefore, when the total cooling load demand changes, it is necessary to adjust the number of operating chillers and their load rates in real time based on the performance characteristics of each chiller unit to ensure high-efficiency operation.

[0003] Currently, most air conditioning and refrigeration room control algorithms for chiller units use performance coefficient (CCF) calculation formulas or tables provided by the chiller manufacturers during pre-shipment testing. However, this approach has several drawbacks: firstly, the CCF calculation formulas or tables only consider the impact of load rate and cooling water temperature on efficiency under rated cooling conditions; secondly, during long-term operation, the performance of chillers will deviate due to factors such as heat exchanger scaling and compressor efficiency. Furthermore, some manufacturers fail to provide performance data at the time of shipment, resulting in poor accuracy and applicability of this method. Additionally, some algorithms use machine learning to build CCF models based on historical operating data. This method is complex, has poor scalability, and the model is difficult to update in real time. Corresponding load allocation optimization methods, such as particle swarm optimization, suffer from high computational complexity, uncontrollable optimization results, and difficulty in intuitive understanding by operating engineers.

[0004] Therefore, those skilled in the art urgently need to find a new method to solve the above problems. Summary of the Invention

[0005] To overcome the problems existing in related technologies, this invention discloses an adaptive optimization control method for a chiller system, the method comprising:

[0006] In a computer room system containing at least two chiller units, a chiller unit performance monitoring relationship table is established and dynamically updated based on real-time collected operating data.

[0007] Based on the dynamic performance monitoring relationship table of the chiller units, determine the chiller unit load distribution operation scheme that meets the total cooling load demand and has the highest system performance coefficient.

[0008] Optionally, the chiller unit performance monitoring relationship table includes four fields: operating condition code, unit number, load rate, and coefficient of performance (COP).

[0009] Optionally, determine the data value interval ΔY for each operation monitoring parameter, encode the values ​​of the operation monitoring parameters, and each code i covers... Value range; all operation monitoring parameter codes are arranged in sequence to form the operation condition code; the real-time collected operation data is mapped to the corresponding operation condition code.

[0010] Optional, the operating monitoring parameters include chilled water inlet temperature, chilled water flow rate, cooling water inlet temperature, and cooling water flow rate.

[0011] Optionally, the method for dynamically updating the performance monitoring relationship table of chiller units includes the following steps:

[0012] Check whether the performance monitoring relationship table already contains the data records corresponding to the operating condition code, unit number, and load rate collected this time;

[0013] If the data record collected this time does not exist in the performance monitoring relationship table, then the data record collected this time will be inserted into the performance monitoring relationship table;

[0014] If the data record collected this time already exists in the performance monitoring relationship table, then check whether the corresponding performance coefficient in the performance monitoring relationship table is the same as the performance coefficient collected this time.

[0015] If they are different, the performance coefficients recorded in the performance monitoring relationship table will be deleted, and the performance coefficients collected this time will be written into the corresponding positions in the performance monitoring relationship table.

[0016] Optionally, when a chiller unit is taken out of operation, the data record corresponding to the unit number of that chiller unit in the chiller unit performance monitoring relationship table is filtered out and deleted.

[0017] Optionally, when a new chiller unit is added to the computer room system, the unit number is determined, and the chiller unit performance monitoring table data records are manually added based on the performance data or experience values ​​provided by the manufacturer.

[0018] Optionally, determining the chiller load allocation operation scheme that meets the total cooling load demand and has the highest system performance coefficient based on the chiller dynamic performance monitoring relationship table includes the following steps:

[0019] Determine the total cooling load demand and the current operating condition code;

[0020] Filter out the data records corresponding to the current operating condition codes in the chiller unit performance monitoring relationship table;

[0021] If the current operating condition does not exist, then filter out the data records of the adjacent operating conditions;

[0022] Based on the selected data records, check whether the load rate range of each unit number covers all the value points. If there are missing value points, add the data record corresponding to the missing value point. The method for adding the data record corresponding to the missing value point is to take the weighted average of the performance coefficients of the two adjacent points above and below as the performance coefficient corresponding to the missing value point.

[0023] An exhaustive method is used to combine all load rate values ​​of each unit to determine the sum of the operating load values ​​of all chiller units in each chiller unit combination scheme. The absolute value of the difference between the sum and the total cooling load demand value is obtained. The chiller unit combination scheme with the absolute value of the difference less than a preset threshold is determined as the chiller unit operation scheme.

[0024] Calculate the system performance coefficient for each chiller unit operation scheme based on the performance coefficient corresponding to each chiller unit.

[0025] The operating scheme of the chiller unit with the highest performance coefficient of the system is taken as the optimal operating scheme.

[0026] Optionally, when using the exhaustive method to combine all load factor values ​​for each unit, if n-1 units have been exhausted, the load factor value of the nth unit can be obtained through the constraint of total cooling load demand.

[0027] Optionally, when using an exhaustive method to combine all load factor values ​​for each unit, if the load factor values ​​for each unit are exhausted in ascending order, the increase in load factor can be stopped when the total load value of all units exceeds the total cooling load demand; if the load factor values ​​are exhausted in descending order, the decrease in load factor can be stopped when the total load value of all units is less than the total cooling load demand.

[0028] Optionally, determining the chiller load allocation operation scheme that meets the total cooling load demand and has the highest system performance coefficient based on the chiller dynamic performance monitoring relationship table includes the following steps:

[0029] Determine the total cooling load demand and the current operating condition code;

[0030] Filter out the data records corresponding to the current operating condition codes in the chiller unit performance monitoring relationship table;

[0031] If the current operating condition does not exist, then filter out the data records of the adjacent operating conditions;

[0032] Based on the selected data records, check whether the load rate range of each unit number covers all the value points. If there are missing value points, add the data record corresponding to the missing value point. The method for adding the data record corresponding to the missing value point is to take the weighted average of the performance coefficients of the two adjacent points above and below as the performance coefficient corresponding to the missing value point.

[0033] Sort the completed data records according to their performance coefficients from highest to lowest;

[0034] Obtain the first data record from the sorted data and record the unit number and the corresponding cooling load value of the load rate into the combination scheme;

[0035] Determine the absolute value of the difference between the cooling load value corresponding to the unit number load rate in the data record and the total cooling load demand value;

[0036] If the absolute value of the difference is greater than a preset threshold, then continue traversing downwards in the sorted data;

[0037] If the difference is negative, and the unit number in the data record retrieved downwards has not yet been recorded in the combination scheme, then the record is retrieved and recorded in the combination scheme;

[0038] If the difference is negative, and the unit number in the data record retrieved downwards has been recorded in the combination scheme, if the load rate in the data record retrieved at this time is higher than the record already recorded in the combination scheme, then the record in the combination scheme will be replaced with the current data record.

[0039] Extract each data record sequentially until the absolute value of the deviation between the cumulative load value and the total cooling load demand value is less than the set threshold.

[0040] Record the unit number corresponding to the last record retrieved, and continue searching downwards for the next record with an increased load rate for other unit numbers. If no record is found, abandon the search.

[0041] If it exists, replace the corresponding unit load rate in the combination scheme to form multiple combinations that meet the total cooling load demand requirements;

[0042] Calculate the system performance coefficients for each combination scheme, and the combination with the highest system performance coefficient is the optimal combination.

[0043] Optionally, during the traversal, if a positive difference is found, the search returns to the first record and continues downwards. All units that have been retrieved in the combination scheme are replaced with records with lower load rates, thus forming multiple combinations that meet the total cooling load requirements. The combination with the highest system performance coefficient is the optimal combination.

[0044] In summary, this invention discloses an adaptive optimization control method for a chiller system. The method includes: in a computer room system containing at least two chillers, establishing and dynamically updating a chiller performance monitoring relationship table based on real-time collected operating data; and determining a chiller load allocation operation scheme that meets the total cooling load demand and has the highest system performance coefficient based on the chiller dynamic performance monitoring relationship table. This method can incorporate the influence of operating parameters such as chilled water inlet temperature, chilled water flow rate, cooling water inlet temperature, and cooling water flow rate, and can adapt to situations where chiller performance parameters are unknown or deviate. It offers high accuracy and reliability; adaptive adjustments can be made when the number of chillers changes; it is adaptable to different computer rooms and has strong versatility.

[0045] Other features and advantages disclosed in this invention will be described in detail in the following detailed description section. Attached Figure Description

[0046] The accompanying drawings are provided to further illustrate the present disclosure and form part of the specification. They are used together with the following detailed description to explain the present disclosure, but do not constitute a limitation thereof. In the drawings:

[0047] Figure 1 This is a flowchart illustrating an adaptive optimization control method for a chiller system according to an exemplary embodiment;

[0048] Figure 2 It is based on Figure 1 The flowchart shown is a method for establishing a performance monitoring relationship table;

[0049] Figure 3 A flowchart of a method for optimizing the operation of a chiller unit;

[0050] Figure 4 This is a flowchart of another method for optimizing the operation of a chiller unit. Detailed Implementation

[0051] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present disclosure.

[0052] Figure 1 This is a flowchart illustrating an adaptive optimization control method for a chiller system according to an exemplary embodiment, such as... Figure 1 As shown, the method includes:

[0053] In step 110, in a computer room system containing at least two chiller units, a chiller unit performance monitoring relationship table is established and dynamically updated based on real-time collected operating data.

[0054] For example, addressing the issue that existing air conditioning chiller room control algorithms neglect the influence of operating data such as chilled water inlet temperature, chilled water flow rate, cooling water inlet temperature, and cooling water flow rate, the embodiments disclosed in this invention organize and optimize the operating condition parameters, unit load rate, and coefficient of performance (COP) data generated during chiller unit operation, establish a chiller unit performance monitoring relationship table, and dynamically update it. In this way, a chiller unit operation combination scheme suitable for the total cooling load demand and with the optimal system COP can be obtained based on this performance monitoring relationship table.

[0055] It is understood that a chiller system includes at least two chiller units, each with different rated load values, operating parameters, and coefficients of performance (COPs). After obtaining the total cooling load demand, to meet this demand, the cooling load needs to be allocated among the chiller units. This means assigning an operating load value to each chiller unit so that the sum of their operating load values ​​reaches the total cooling load demand. Furthermore, different operating load values ​​for each chiller unit will result in different COPs. The embodiments disclosed in this invention seek a chiller unit operating combination scheme with the highest COP. It is understood that the COP is the ratio of the total system load to the total system power consumption in the chiller unit combination scheme.

[0056] Specifically, the performance monitoring relationship table for the chiller unit includes four fields: operating condition code, unit number, load rate, and coefficient of performance (COP).

[0057] The operating condition coding in the performance monitoring relationship involves the following steps: determining the data value interval ΔY for each operating monitoring parameter, coding the values ​​of the operating monitoring parameters, with each code i covering... Value range; all operation monitoring parameter codes are arranged in sequence to form the operation condition code; the real-time collected operation data is mapped to the corresponding operation condition code.

[0058] It should be noted that during the operation of the chiller unit, various operational data are collected in real time to form a database. For the data in this database, the data value interval ΔY is determined, and the operational monitoring parameters (including chilled water inlet temperature, chilled water flow rate, cooling water inlet temperature, and cooling water flow rate) are coded. All operational monitoring parameter codes are sequentially arranged to form an operational condition code; the real-time collected operational data is mapped to the corresponding operational condition code in the performance monitoring relationship table. It is understandable that the data value interval ΔY needs to be determined based on the sensor accuracy and the influence of the operational monitoring parameters on the performance coefficient. For example, the higher the sensor accuracy, the greater the influence of the operational monitoring parameters on the performance coefficient; therefore, the smaller the corresponding data value interval ΔY should be set, and the more densely the operational monitoring parameter values ​​obtained based on the data value interval ΔY.

[0059] To facilitate the classification and organization of load factor and performance coefficient based on the above four types of operational monitoring parameters, the operational monitoring parameters are encoded in the embodiments disclosed in this invention to obtain the operational condition code Y. i For example, in the operation monitoring parameter code N01020304, position 01 represents the chilled water inlet temperature code, and the value 01 indicates that the chilled water inlet temperature code is 01. Position 02 represents the chilled water flow rate code, and the value 02 indicates that the chilled water flow rate code is 02. Position 03 represents the cooling water inlet temperature code, and the value 03 indicates that the cooling water inlet temperature code is 03. Position 04 represents the cooling water flow rate code, and the value 04 indicates that the cooling water flow rate code is 04. The values ​​of different chilled water inlet temperatures, chilled water flow rates, cooling water inlet temperatures, and cooling water flow rates are combined and compiled into an operation condition code, and the unit load rate and performance coefficient under that operation condition code are recorded accordingly.

[0060] It is understandable that when the data value interval is ΔY, the parameter range covered by each operating condition code is...

[0061] For example, taking the chilled water inlet temperature as an example (assuming the lowest operating temperature of chilled water is 3℃, the highest operating temperature is 12℃, and the temperature interval is 1℃), the data sampling situation is shown in Table 1:

[0062] Table 1

[0063]

[0064] The performance monitoring relationships of the chiller units (assuming the unit load rate values ​​are taken at intervals of 0.1) are shown in Table 2:

[0065] Table 2

[0066] <![CDATA[ Operating condition coding ]]> <![CDATA[ Unit number ]]> <![CDATA[ Load factor ]]> COP N01010101 1 1 5.5 N01010101 1 0.9 6.0 … … … … N09020304 8 0.3 6.0 … … … …

[0067] Specifically, Figure 2 It is based on Figure 1 The flowchart shown is a method for dynamically updating a performance monitoring relationship table, such as... Figure 2 As shown, this step includes:

[0068] In step 210, it is checked whether the performance monitoring relationship table already contains the data records corresponding to the operating condition code, unit number, and load rate collected this time.

[0069] For example, in the disclosed embodiments of the present invention, while establishing the performance monitoring relationship table, it is also necessary to continuously optimize and update the performance monitoring relationship table based on the operating condition parameters, unit load rate and performance coefficient collected in real time during unit operation.

[0070] It should be noted that, based on the data value intervals, the data records corresponding to the operating condition code, unit number, and load rate are first determined to be valid. Only if the data records are valid can the performance monitoring relationship table be optimized and updated using that data and these records. Specifically, it is determined whether the operating condition parameters, unit load rate, and Coefficient of Performance (COP) have changed. If any of these data changes, it is determined whether the data change is stable, thereby determining whether the operating condition is stable. It can be understood that determining whether the data change is stable can be based on historical operating data. Once stable operating condition is confirmed, the data record is then determined to be valid. If the preset sensors within the chiller system collect abnormal data, causing an alarm message to be issued by the chiller room, the operating condition code and the corresponding unit load rate and COP collected this time are determined to be invalid data; if no alarm message is issued, the data is considered valid.

[0071] For valid data records, determine whether the valid data records acquired this time already exist in the performance monitoring relationship table, and determine whether the data records acquired this time need to be inserted into the performance monitoring relationship table according to the following steps 220-240, so as to complete the update of the performance monitoring relationship table.

[0072] In step 220, if the data record collected this time does not exist in the performance monitoring relationship table, then the data record collected this time will be inserted into the performance monitoring relationship table.

[0073] In step 230, if the data record collected this time already exists in the performance monitoring relationship table, then it is checked whether the corresponding performance coefficient in the performance monitoring relationship table is the same as the performance coefficient collected this time.

[0074] In step 240, if they are not the same, the performance coefficient recorded in the performance monitoring relationship table is deleted, and the performance coefficient collected this time is written into the corresponding position in the performance monitoring relationship table.

[0075] For example, the system checks if the operating condition code for this group already exists in the performance monitoring relationship table. If it does, it checks if the performance coefficient corresponding to the operating condition code in the performance monitoring relationship table matches the data collected this time. If they don't match, the collected performance coefficient is written to the corresponding position of the same operating condition code in the performance monitoring relationship table. If they match, no modification to the data in the performance monitoring relationship table is needed. Additionally, it can be understood that if a valid operating condition code, unit number, load factor, and performance coefficient (COP) are collected, but the corresponding operating condition code is not recorded in the performance monitoring relationship table, then the collected operating condition code, along with the corresponding unit load factor and performance coefficient, can be directly inserted into the corresponding positions in the performance monitoring relationship table.

[0076] In step 250, when a chiller unit is taken out of operation, the data record corresponding to the unit number of the chiller unit in the chiller unit performance monitoring relationship table is filtered out and deleted.

[0077] In step 260, when a new chiller unit is added to the computer room system, the unit number is determined, and the chiller unit performance monitoring table data is manually added according to the performance data or experience values ​​provided by the manufacturer.

[0078] For example, regarding the data records in the aforementioned performance monitoring relationship table, when a chiller unit is taken out of operation, the corresponding data record for that chiller unit is deleted from both the chiller unit parameter relationship table and the chiller unit performance monitoring relationship table. When a new chiller unit is added, a data record is added to the chiller unit parameter relationship table. Furthermore, during initial operation, data records in the chiller unit performance monitoring table can be manually added based on the performance data or empirical values ​​provided by the manufacturer.

[0079] In step 120, based on the dynamic performance monitoring relationship table of the chiller units, a load distribution operation scheme for the chiller units that meets the total cooling load demand and has the highest system performance coefficient is determined.

[0080] For example, a performance monitoring relationship table is created based on the operating condition codes, unit numbers, load rates, and coefficient of performance (COP) collected in the above steps. This table reflects the operating condition parameters of each unit and the relationship between the unit load rate and the COP. Thus, once the total cooling load demand of the chiller system is determined, the optimal operating scheme that meets the total cooling load demand and has the highest COP can be determined based on this performance monitoring relationship table. It is also important to note that the operating load value of each chiller unit in the scheme must be less than its rated load value (i.e., the chiller units cannot be overloaded).

[0081] For example, such as Figure 3 and Figure 4 The image shows two embodiments of the present invention for querying and optimizing performance monitoring relationship tables to obtain chiller unit operation schemes.

[0082] Before introducing the following two disclosed embodiments, it should be clarified that, for ease of explanation, both embodiments described below use three units as examples. Units 1 and 2 have a rated cooling capacity of 4000kW, and unit 3 has a rated cooling capacity of 2000kW. The unit load rate values ​​are spaced at intervals of 0.1. Furthermore, the total cooling load demand is set to 8000kW. This embodiment is for illustrative purposes only and is not intended for actual operation.

[0083] Figure 3 A flowchart of a method for optimizing the operation of a chiller unit, such as... Figure 3 As shown in the table, based on the dynamic performance monitoring relationship table of the chiller units, the load distribution operation scheme of the chiller units that meets the total cooling load demand and has the highest system performance coefficient is determined, including the following steps:

[0084] In step 310, the total cooling load demand value and the current operating condition code are determined.

[0085] In step 320, the data records corresponding to the current operating condition codes in the chiller unit performance monitoring relationship table are selected.

[0086] In step 330, if the current operating condition does not exist, the data records of the adjacent operating conditions are filtered out.

[0087] For example, when it is necessary to allocate the load of each chiller unit according to the total cooling load demand so that the sum of the operating load values ​​of each chiller unit reaches the total cooling load demand, it is first necessary to determine the total cooling load demand and the current operating condition code. Based on the current operating condition code, the data records for the current operating condition are filtered out from the performance monitoring relationship table. It should be noted that if the current operating condition code does not exist in the performance monitoring relationship table, adjacent operating condition data records are filtered out. Adjacent operating condition data records refer to the group of data records in the performance monitoring relationship table whose values ​​are closest to the current operating condition code.

[0088] In step 340, based on the filtered data records, it is checked whether the load rate value range of each unit number covers all the value points. If there are missing value points, the data record corresponding to the missing value point is added. The method for adding the data record corresponding to the missing value point is to take the weighted average of the performance coefficients of the two adjacent points above and below as the performance coefficient corresponding to the missing value point.

[0089] For example, before querying the optimal chiller unit operation scheme, for the selected data records, it is also necessary to check whether the load rate value range corresponding to each unit number can cover all the value points according to the performance monitoring relationship table; check whether the load rate value range of each unit number covers all the value points. If a value point is missing, add the record of that value point, and take the weighted average of the two adjacent points above and below. The weight is determined according to the distance between the value point and the two adjacent points above and below.

[0090] Specifically, the data records for a specific operating condition in the chiller unit performance monitoring relationship table are shown in Table 3:

[0091] Table 3

[0092]

[0093]

[0094] As shown in Table 3, if the data set for unit number 1, unit load rate 0.7, and COP 6.7 is missing, it indicates that the unit load rate does not cover all value points. In this case, the weighted average of the two adjacent data sets (i.e., unit number 1, unit load rate 0.8, COP 6.4 and unit number 1, unit load rate 0.6, COP 6.8) is calculated (the weight is determined based on the distance between this value point and its two adjacent points above and below), thus obtaining the COP value corresponding to the unit load rate of 0.7. It can be understood that, when determining whether the unit load rate value points are missing, one possible method in the embodiments of this invention is to first sort all the unit load rates corresponding to each chiller unit in ascending or descending order (as shown in Table 3, descending order, i.e., sorting the unit load rates from largest to smallest). After sorting, a data set of unit load rates that gradually increase or decrease in an arithmetic sequence is obtained, and the presence of missing data is determined according to the tolerance. Similarly, for a missing value location, the load rates of the two adjacent units are the load rates of the units whose values ​​are closest to the missing value in the arithmetic sequence. Determining the load rates of the two adjacent units and calculating their average is equivalent to calculating the load rate of the unit at the missing value location. It is also necessary to calculate the weighted average of the COP values ​​corresponding to the load rates of the two adjacent units to calculate the COP value at the missing value location.

[0095] In step 350, an exhaustive method is used to combine all load rate values ​​of each unit to determine the sum of the operating load values ​​of all chiller units in each chiller unit combination scheme. The absolute value of the difference between the sum and the total cooling load demand value is obtained, and the chiller unit combination scheme with the absolute value of the difference less than a preset threshold is determined as the chiller unit operation scheme.

[0096] In step 360, the system performance coefficient of each chiller unit operation scheme is calculated based on the performance coefficient corresponding to each chiller unit.

[0097] In step 370, the operating scheme of the chiller unit with the highest performance coefficient of the system is taken as the optimal operating scheme.

[0098] For example, in the disclosed embodiment of the present invention, an exhaustive method is used to combine and sum all operating load values ​​(obtained from the unit load rate) of each unit to obtain a sum. Among the combined schemes, combinations whose absolute deviation between the sum and the total cooling load demand is less than a set threshold are searched and selected as the chiller unit operating schemes. The system performance coefficient of each unit in the chiller unit operating schemes is calculated. The combination with the highest system performance coefficient is the optimal combination.

[0099] For example, all combinations that meet the total cooling load demand are shown in the table below. The optimal combination has a combined COP of 6.62. If a conventional control strategy is adopted (units 1 and 2 are turned on), the combined COP is only 5.65. The optimized COP is improved by 10.9%. Specifically, as shown in Table 4:

[0100] Table 4

[0101]

[0102] Furthermore, when using the exhaustive method to combine all load factor values ​​for each unit, if n-1 units have been exhausted, the load factor value of the nth unit can be obtained through the constraint of the total cooling load demand value. In other words, if n-1 units have been exhausted, in order to reduce the amount of calculation, the nth unit can be selected from the chillers above and below based on the constraint of the total cooling load demand value.

[0103] Understandably, when using the exhaustive method to combine all load factor values ​​for each unit, if the load factor values ​​for each unit are exhausted in ascending order, the process of increasing the load factor can be stopped when the total load value of all units exceeds the total cooling load demand; if the load factor values ​​are exhausted in descending order, the process of decreasing the load factor can be stopped when the total load value of all units is less than the total cooling load demand.

[0104] Figure 4 A flowchart for another method of optimizing the operation of chiller units, such as Figure 4 As shown, the chiller load allocation operation scheme that meets the total cooling load demand and has the highest system performance coefficient, based on the dynamic performance monitoring relationship table of the chiller units, includes the following steps:

[0105] In step 401, the total cooling load demand value and the current operating condition code are determined.

[0106] In step 402, the data records corresponding to the current operating condition codes in the chiller unit performance monitoring relationship table are selected.

[0107] In step 403, if the current operating condition does not exist, the data records of the adjacent operating conditions are filtered out.

[0108] In step 404, based on the filtered data records, it is checked whether the load rate value range of each unit number covers all value points. If there are missing value points, the data record corresponding to the missing value point is added.

[0109] The method for adding the data record corresponding to the missing value is as follows: take the weighted average of the performance coefficients of the two adjacent points above and below, and use it as the performance coefficient corresponding to the missing value.

[0110] For example, the steps of determining the total cooling load demand value and the current operating condition code, as well as filtering data records and adding data records corresponding to missing value points, are the same as the principles of steps 310-340 above, and will not be repeated in this embodiment of the invention.

[0111] In step 405, the supplemented data records are sorted from high to low according to their performance coefficients.

[0112] In step 406, the first data record in the sorted data is obtained, and the unit number and the corresponding load rate cooling load value are recorded in the combination scheme.

[0113] In step 407, the absolute value of the difference between the cooling load value corresponding to the unit number load rate in the data record and the total cooling load demand value is determined.

[0114] In step 408, if the absolute value of the difference is greater than a preset threshold, the process continues to traverse downwards in the sorted data.

[0115] In step 409, if the difference is negative and the unit number in the data record retrieved downwards has not yet been recorded in the combination scheme, then the record is retrieved and recorded in the combination scheme.

[0116] In step 410, if the difference is negative and the unit number in the data record retrieved downwards has been recorded in the combination scheme, and if the load rate in the data record retrieved at this time is higher than the record already recorded in the combination scheme, then the record in the combination scheme will be replaced with the current data record.

[0117] In step 411, each data record is retrieved sequentially until the absolute value of the deviation between the cumulative load value and the total cooling load demand value is less than the set threshold.

[0118] In step 412, record the unit number corresponding to the last record retrieved, and continue searching downwards for the next record with an increased load rate for other unit numbers. If no record is found, discard the record.

[0119] In step 423, if the load rate of the corresponding unit in the combination scheme is replaced, multiple combinations that meet the total cooling load demand requirements are formed.

[0120] In step 414, the system performance coefficients of each combination scheme are calculated, and the combination with the highest system performance coefficient is the optimal combination.

[0121] For example, the current operating condition data records are sorted in descending order of performance coefficient. If the performance coefficients are the same, they are sorted by unit number and unit load rate respectively. First, the first data record is retrieved, and the absolute value of the difference between it and the total cooling load demand is calculated. If it is greater than a set threshold, the search continues downward. If the difference is negative, and the unit number in the retrieved data records has not yet been recorded and accumulated, the record is retrieved and added to the combination scheme. Similarly, if the difference is negative, for the retrieved and already recorded unit combination schemes, if the load rate of the retrieved unit combination scheme is higher than the record of the already recorded combination scheme, the record in the combination scheme is replaced with the current data value. For example, the load difference value corresponding to the first data record taken from Table 6 is -5600. For the unit combination schemes taken down and already recorded (the second row in Table 6), if the load rate of the taken unit combination scheme is higher than that of the already recorded combination scheme (the load value of Unit 2 in the second row of data records is 2800, which is higher than the load value of Unit 2 in the first row of data records, indicating that the load rate in the second row of data records is higher than that in the first row of data records), then the record in the combination scheme will be replaced with the current data value.

[0122] Using the above method, each data record is retrieved sequentially until the absolute value of the difference between the sum of the operating load values ​​and the total cooling load demand is less than a set threshold. The unit number corresponding to the last retrieved data record is recorded. The search continues downwards for the next record with an increased unit load rate for other unit numbers. If no record is found, it is discarded; if a record exists, the corresponding unit load rate in the combination scheme is replaced, forming multiple combinations that meet the total cooling load demand requirements. The corresponding comprehensive COP (Coefficient of Performance) is calculated, and the combination with the highest comprehensive COP is the optimal combination.

[0123] In addition, if a positive difference is found during the traversal, the search returns to the first record and continues downwards. All the units that have been retrieved in the combination scheme are replaced with records with lower load rates, thus forming multiple combinations that meet the total cooling load requirements. The combination with the highest system performance coefficient is the optimal combination.

[0124] Specifically, the data records of the selected chiller unit performance monitoring relationship table are arranged from high to low according to the coefficient of performance (COP), as shown in Table 5.

[0125] Table 5

[0126]

[0127] The optimization process is shown in Table 6. In this example, unit number 3 is already at full load, so two combinations that meet the requirements are ultimately formed. The corresponding integrated COP is calculated, and the combination with the higher integrated COP is the optimal combination.

[0128] Table 6

[0129]

[0130]

[0131] In addition, when a chiller unit is taken out of operation, the corresponding data record for that chiller unit will be deleted from the chiller unit parameter relationship table and the chiller unit performance monitoring relationship table. When a new chiller unit is added, a data record will be added to the chiller unit parameter relationship table; during initial operation, data records in the chiller unit performance monitoring table can be manually added based on the performance data or experience values ​​provided by the manufacturer.

[0132] In summary, this invention discloses an adaptive optimization control method for a chiller system. The method includes: in a computer room system containing at least two chillers, establishing and dynamically updating a chiller performance monitoring relationship table based on real-time collected operating data; and determining a chiller load allocation operation scheme that meets the total cooling load demand and has the highest system performance coefficient based on the chiller dynamic performance monitoring relationship table. This method can incorporate the influence of operating parameters such as chilled water inlet temperature, chilled water flow rate, cooling water inlet temperature, and cooling water flow rate, and can adapt to situations where chiller performance parameters are unknown or deviate. It offers high accuracy and reliability; adaptive adjustments can be made when the number of chillers changes; it is adaptable to different computer rooms and has strong versatility.

[0133] The preferred embodiments of this disclosure have been described in detail above with reference to the accompanying drawings. However, this disclosure is not limited to the specific details of the above embodiments. Within the scope of the technical concept of this disclosure, various simple modifications can be made to the technical solutions of this disclosure, and these simple modifications all fall within the protection scope of this disclosure.

[0134] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, this disclosure will not describe the various possible combinations separately.

[0135] Furthermore, various different embodiments of this disclosure can be combined in any way, as long as they do not violate the spirit of this disclosure, they should also be regarded as the content disclosed in this disclosure.

Claims

1. An adaptive optimization control method for a chiller unit system, characterized in that, The method includes: In a computer room system containing at least two chiller units, a chiller unit performance monitoring relationship table is established and dynamically updated based on real-time collected operating data. Based on the dynamic performance monitoring relationship table of the chiller units, determine the load distribution operation scheme of the chiller units that meets the total cooling load demand and has the highest system performance coefficient. The chiller unit performance monitoring relationship table includes four fields: operating condition code, unit number, load rate, and performance coefficient. Operational monitoring parameters include chilled water inlet temperature, chilled water flow rate, cooling water inlet temperature, and cooling water flow rate; Determine the data range for each operational monitoring parameter. The values ​​of the operation monitoring parameters are encoded, and each code... cover Range of values; where, The center value of the range of values ​​for the operation monitoring parameters covered by each code i; all operation monitoring parameter codes are arranged in sequence to form the operation condition code; The real-time collected operating data is mapped to the corresponding operating condition code.

2. The adaptive optimization control method for a chiller system according to claim 1, characterized in that, The method for dynamically updating the performance monitoring relationship table of chiller units includes the following steps: Check whether the performance monitoring relationship table already contains the data records corresponding to the operating condition code, unit number, and load rate collected this time; If the data record collected this time does not exist in the performance monitoring relationship table, then the data record collected this time will be inserted into the performance monitoring relationship table; If the data record collected this time already exists in the performance monitoring relationship table, then check whether the corresponding performance coefficient in the performance monitoring relationship table is the same as the performance coefficient collected this time. If they are different, the performance coefficients recorded in the performance monitoring relationship table will be deleted, and the performance coefficients collected this time will be written into the corresponding positions in the performance monitoring relationship table.

3. The adaptive optimization control method for a chiller system according to claim 1, characterized in that, The process of determining the chiller load allocation operation scheme that meets the total cooling load demand and has the highest system performance coefficient based on the chiller dynamic performance monitoring relationship table includes the following steps: Determine the total cooling load demand and the current operating condition code; Filter out the data records corresponding to the current operating condition codes in the chiller unit performance monitoring relationship table; If the current operating condition does not exist, then filter out the data records of the adjacent operating conditions; Based on the selected data records, check whether the load rate range of each unit number covers all the value points. If there are missing value points, add the data record corresponding to the missing value point. The method for adding the data record corresponding to the missing value point is to take the weighted average of the performance coefficients of the two adjacent points above and below as the performance coefficient corresponding to the missing value point. An exhaustive method is used to combine all load rate values ​​of each unit to determine the sum of the operating load values ​​of all chiller units in each chiller unit combination scheme. The absolute value of the difference between the sum and the total cooling load demand value is obtained. The chiller unit combination scheme with the absolute value of the difference less than a preset threshold is determined as the chiller unit operation scheme. Calculate the system performance coefficient for each chiller unit operation scheme based on the performance coefficient corresponding to each chiller unit. The operating scheme of the chiller unit with the highest performance coefficient of the system is taken as the optimal operating scheme.

4. The method according to claim 3, characterized in that: When using the exhaustive method to combine all load factor values ​​for each unit, if n-1 units have been exhausted, the load factor value of the nth unit can be obtained by using the total cooling load demand constraint.

5. The method according to claim 4, characterized in that: When using the exhaustive method to combine all load factor values ​​for each unit, if the load factor values ​​for each unit are exhausted in ascending order, the increase in load factor value can be stopped when the total load value of all units exceeds the total cooling load demand value; if the load factor values ​​are exhausted in descending order, the decrease in load factor value can be stopped when the total load value of all units is less than the total cooling load demand value.

6. The adaptive optimization control method for a chiller system according to claim 1, characterized in that, The process of determining the chiller load allocation operation scheme that meets the total cooling load demand and has the highest system performance coefficient based on the chiller dynamic performance monitoring relationship table includes the following steps: Determine the total cooling load demand and the current operating condition code; Filter out the data records corresponding to the current operating condition codes in the chiller unit performance monitoring relationship table; If the current operating condition does not exist, then filter out the data records of the adjacent operating conditions; Based on the selected data records, check whether the load rate range of each unit number covers all the value points. If there are missing value points, add the data record corresponding to the missing value point. The method for adding the data record corresponding to the missing value point is to take the weighted average of the performance coefficients of the two adjacent points above and below as the performance coefficient corresponding to the missing value point. Sort the completed data records according to their performance coefficients from highest to lowest; Obtain the first data record from the sorted data and record the unit number and the corresponding cooling load value of the load rate into the combination scheme; Determine the absolute value of the difference between the cooling load value corresponding to the unit number load rate in the data record and the total cooling load demand value; If the absolute value of the difference is greater than a preset threshold, then continue traversing downwards in the sorted data; If the difference is negative, and the unit number in the data record retrieved downwards has not yet been recorded in the combination scheme, then the record is retrieved and recorded in the combination scheme; If the difference is negative, and the unit number in the data record retrieved downwards has been recorded in the combination scheme, if the load rate in the data record retrieved at this time is higher than the record already recorded in the combination scheme, then the record in the combination scheme will be replaced with the current data record. Extract each data record sequentially until the absolute value of the deviation between the cumulative load value and the total cooling load demand value is less than the set threshold. Record the unit number corresponding to the last record retrieved, and continue searching downwards for the next record with an increased load rate for other unit numbers. If no record is found, abandon the search. If it exists, replace the corresponding unit load rate in the combination scheme to form multiple combinations that meet the total cooling load demand requirements; Calculate the system performance coefficients for each combination scheme, and the combination with the highest system performance coefficient is the optimal combination.

7. The method according to claim 6, characterized in that, During the traversal, if a positive difference is found, the search returns to the first record and continues downwards. All units that have been retrieved in the combination scheme are replaced with records of lower load rates, thus forming multiple combinations that meet the total cooling load requirements. The combination with the highest system performance coefficient is the optimal combination.

Images