A whole-process intelligent monitoring and energy efficiency management method for a ground source heat pump system

Abstract

The application relates to the technical field of ground source heat pump control, and discloses a full-process intelligent monitoring and energy efficiency management method for a ground source heat pump system. First, a hardware-level digital signature is generated by using a physically unclonable function unit at a data acquisition end. Then, double verification is performed, including signature-based security constraint verification and physical constraint verification based on a physical constraint set. According to the verification result, the data flow is classified into trusted data flow, invalid data flow and security alarm flow. The trusted data flow is used for real-time working condition adjustment and energy efficiency control. Meanwhile, the trusted data flow is used for long-time window cumulative analysis to analyze the long-time change of the physical characteristics of the system. When the cumulative analysis result of the model measurement deviation meets a preset condition, a dynamic physical constraint is generated, and the dynamic physical constraint is used for feedback updating of the physical constraint set. Through double verification and dynamic constraint evolution, the safety reliability and long-time operation energy efficiency of the system are improved.

Description

Technical Field

[0001] This invention relates to the field of ground source heat pump control technology, specifically to a method for intelligent monitoring and energy efficiency management of the entire process of a ground source heat pump system. Background Technology

[0002] As a highly efficient building energy-saving technology, the operating energy efficiency of ground source heat pump systems (GSHP) largely depends on advanced monitoring and control systems.

[0003] Existing monitoring systems generally suffer from insufficient emphasis on data reliability. Traditional control systems typically assume that data collected by sensors is authentic and accurate, lacking effective authentication mechanisms for data sources. This makes the system vulnerable to cyberattacks such as data tampering and forgery. Furthermore, the system lacks robust identification and processing capabilities for physical faults, transient failures, or signal drift occurring within the sensors themselves. Once this unreliable or faulty data enters the control logic, it can lead to incorrect operating condition adjustments, causing system instability or increased energy consumption.

[0004] Furthermore, ground source heat pumps are systems whose physical characteristics change slowly over time. For example, buried pipe heat exchangers or plate heat exchangers experience a gradual decrease in heat transfer efficiency due to scaling, and equipment such as water pumps also suffer wear and tear. However, existing energy efficiency management methods largely rely on static physical models or control parameters set during the initial system deployment. These static models cannot reflect the physical aging of the system due to long-term operation. As the actual physical characteristics deviate from the initial design values, the original control strategies will no longer be optimal, leading to a significant decline in system energy efficiency over extended operating time. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method for intelligent monitoring and energy efficiency management of the entire process of a ground source heat pump system. It aims to solve the deficiencies of existing ground source heat pump monitoring systems in terms of data security, physical fault identification, and handling of long-term system performance aging, thereby improving the system's operational reliability and energy efficiency.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for intelligent monitoring and energy efficiency management of the entire process of a ground source heat pump system, wherein the method:

[0007] First, raw operating data is collected at the sensing end of the ground source heat pump system. To ensure the authenticity and integrity of the data source, the method utilizes a physically non-cloning function unit integrated at the data acquisition end to generate a digital signature based on hardware physical characteristics for each data packet.

[0008] The data packet carrying the signature is then sent to a verification node, where two-factor authentication is performed. This two-factor authentication process involves two layers:

[0009] The first step is security constraint verification, which verifies the hardware-level digital signature of the data packet by checking the pre-stored public key, and also checks the timestamp of the data packet to prevent replay attacks. Only data that passes this layer of security verification is confirmed as having a trustworthy source.

[0010] After confirming the reliability of the data source, a second level of physical constraint verification is performed. This verification compares the original operational data within the data packet with a set of physical constraints. This set of physical constraints contains the physical rules that the system should follow, such as static thresholds based on the physical range of the equipment, multivariate correlations based on thermodynamic laws, or timing logic based on the maximum physical rate of change.

[0011] Based on the results of the dual verification, the data streams are categorized and processed. Data packets whose verification results indicate an untrusted data source are classified as security alarm streams and routed to the security response module, which can perform isolation measures such as blacklisting the data source. Data packets with a trustworthy source but whose data content does not conform to physical constraints are classified as failed data streams for the diagnosis and logging of physical faults. Data packets that pass both security and physical constraint verifications are classified as trusted data streams.

[0012] The method of this invention includes a dual-loop control architecture. The first control loop is a real-time control loop that uses the aforementioned reliable data stream. Based on this reliable data, the system solves a real-time optimization problem aimed at minimizing total power consumption, generating optimal control commands to adjust the real-time operating conditions of the ground source heat pump system and achieve energy efficiency optimization.

[0013] The second control loop is a long-term evolutionary loop. This loop aims to enable the system to adapt to long-term changes in physical characteristics caused by factors such as equipment aging and scaling. In this loop, the system compares the measured values ​​in the reliable data stream with the predicted values ​​of a preset physical model, forming a model-measured deviation. By performing cumulative analysis of this deviation over a long window, the system can determine whether the physical characteristics have changed significantly.

[0014] When the accumulated deviation analysis results meet the preset constraint evolution conditions, it indicates that the system's physical characteristics have changed significantly, and the original static physical constraints may no longer be applicable. At this point, the system will initiate a constraint evolution program to generate dynamic physical constraints that reflect the current actual physical characteristics of the system based on recently accumulated reliable operational data. These dynamic physical constraints will be fed back to the nodes performing dual verification to update their physical constraint sets.

[0015] This invention, through the aforementioned technical solution, first establishes a trusted foundation for data sources at the hardware level using physically non-cloning functions. Then, combining a dual verification mechanism of security and physical verification, it filters the data, distinguishing between three types of data streams: security alarms, physical failures, and trusted operation. Based on this, the method's dual-loop control architecture combines real-time energy efficiency control based on trusted data with long-term evolution to address system aging. Through dynamic feedback updates of physical constraints, the system's physical constraint verification criteria can adaptively adjust with long-term changes in the system's physical characteristics, thereby optimizing system operating energy efficiency while improving its operational reliability and security throughout its entire lifecycle.

[0016] This invention provides a method for intelligent monitoring and energy efficiency management of the entire process of a ground source heat pump system. It has the following beneficial effects:

[0017] 1. This invention utilizes physically unclonable function units to generate hardware-level signatures at the data acquisition end and verifies these signatures through security constraint verification. This approach ensures the authenticity of the data source and the integrity of the content from the data origin, effectively preventing data from being tampered with or forged during transmission and guaranteeing that the data foundation upon which subsequent control commands are based is trustworthy.

[0018] 2. This invention classifies data streams through dual verification using both security and physical constraints. In particular, the physical constraint verification identifies data with a reliable source but whose content does not conform to physical laws (such as momentary sensor malfunctions) as failed data streams and prevents them from entering the real-time energy efficiency control loop. This avoids control decision errors caused by abnormal data disturbances, improves the stability of real-time control, and provides a clear basis for physical fault diagnosis.

[0019] 3. The second control loop (long-term evolution loop) of this invention solves the problem of inaccurate monitoring caused by long-term changes in system physical characteristics (such as equipment aging and scaling). This loop accumulates and analyzes the measured deviations of the model, and when it determines that the system characteristics have changed significantly, it generates dynamic physical constraints and updates the physical constraint set accordingly. This adaptive mechanism enables the physical verification standard to evolve with the actual operating conditions of the system, avoiding the misjudgment of normal operating data of an aging system as failed data, and ensuring the long-term effectiveness of reliable data streams. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the system architecture of one embodiment of the present invention;

[0021] Figure 2 This is a schematic diagram of a method flow according to an embodiment of the present invention.

[0022] Among them, 10 is the trusted data acquisition module; 20 is the PCV gateway module; 30 is the dual-loop control execution module; 31 is the real-time energy efficiency control submodule; 32 is the long-term evolution and constraint generation submodule; 40 is the safety response module; and 50 is the system actuator. Detailed Implementation

[0023] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] See attached document Figure 1 This invention provides a method for intelligent monitoring and energy efficiency management of the entire process of a ground source heat pump system, which can be operated in... Figure 1 The system architecture shown can logically include: a trusted data acquisition module 10, a PCV gateway module 20, a dual-loop control execution module 30, and a security response module 40.

[0025] The trusted data acquisition module 10 is located at the physical sensing end of the ground source heat pump system, such as at sensors or their data transmission units (DTUs) deployed in underground pipes, water pumps, main units, or indoor terminals. The trusted data acquisition module 10 integrates a physically unclonable function unit. This physically unclonable function unit is used to provide a unique, hardware-based physical identifier for the sensor.

[0026] The trusted data acquisition module 10 is also used to perform digital signatures on the acquired raw operating data of the ground source heat pump system (such as temperature, flow rate, pressure, valve opening, etc.). The signature is generated using the hardware key provided by the physically non-clonable function unit, thereby forming a data packet carrying the hardware signature.

[0027] PCV gateway module 20 has its data input terminal communicatively connected to trusted data acquisition module 10. PCV gateway module 20 is responsible for receiving the aforementioned data packets carrying signatures and performing dual authentication.

[0028] The dual authentication of the PCV gateway module 20 includes security constraint authentication and physical constraint authentication. Security constraint authentication verifies the signature of the data packet by retrieving the pre-stored L-PUF public key to confirm the authenticity and integrity of the data source.

[0029] After the security constraint verification passes, the PCV gateway module 20 further performs physical constraint verification on the original data content in the data packet. This verification is based on a preset (or dynamically updated) set of physical constraints. This involves determining whether the data content violates the laws of thermodynamics, fluid conservation, or the physical rationality of the system's timing logic.

[0030] PCV gateway module 20 is also used to classify data flows based on the results of dual authentication and to perform heterogeneous data flow routing. Data flows are classified into: trusted data flows that have passed both security and physical authentication; failed data flows that have passed security authentication but failed physical authentication; and security alarm flows that have failed security authentication.

[0031] The dual-loop control execution module 30 has its data input connected to the outputs of the trusted data stream and the failed data stream of the PCV gateway module 20. Internally, the dual-loop control execution module 30 includes a real-time energy efficiency control submodule 31 and a long-term evolution and constraint generation submodule 32.

[0032] The real-time energy efficiency control submodule 31 only receives reliable data streams and performs real-time operating condition adjustment and energy efficiency control of the ground source heat pump system based on the highly reliable data (e.g., adjusting the water pump frequency or compressor load).

[0033] The long-term evolution and constraint generation submodule 32 simultaneously receives trusted data streams (e.g., forwarded from the real-time energy efficiency control submodule 31) and failure data streams. The long-term evolution and constraint generation submodule 32 is primarily responsible for processing the trusted data streams ( The model's measured bias contained in the data is accumulated over a long window for analysis, which is used to perform parameter tuning or generate dynamic physical constraints (DPCs) when specific evolutionary conditions are met; at the same time, this long-term evolution and constraint generation submodule 32 can also process the received failure data stream ( Perform physical fault diagnosis and log recording.

[0034] In this embodiment, the output of the long-term evolution and constraint generation submodule 32 is connected to the physical constraint set of the PCV gateway module 20. The storage unit has a feedback connection. This connection is used to send newly generated dynamic physical constraints back to the PCV gateway module 20, enabling dynamic updates of its physical constraint rules.

[0035] The security response module 40 has its data input terminal connected to the security alarm stream output terminal of the PCV gateway module 20. The security response module 40 is used to trigger a preset security response strategy when it receives a security alarm stream (i.e., the data packet source is untrusted). For example, it may immediately isolate the data source or switch the ground source heat pump system to a preset security degradation operation mode.

[0036] See attached document Figure 2 This invention provides a method for intelligent monitoring and energy efficiency management of the entire process of a ground source heat pump system. This method can be based on... Figure 1The system architecture shown is executed. This method may include the following steps:

[0037] Step S10: Data Acquisition and Signature. The trusted data acquisition module 10 acquires the raw operating data of the ground source heat pump system and uses its integrated physically unclonable function unit to generate a hardware-level digital signature, forming a data packet carrying the signature.

[0038] Step S20: Dual authentication. PCV gateway module 20 receives the data packet. PCV gateway module 20 first performs security constraint verification, that is, verifies the L-PUF signature of the data packet.

[0039] Provided that the security constraint verification is passed, the PCV gateway module 20 further performs physical constraint verification, that is, compares the original data content in the data packet with the preset (or dynamically updated) physical constraint set (R) to determine whether the data content has physical rationality.

[0040] Step S30: Heterogeneous data flow routing. Based on the dual authentication result of step S20, the PCV gateway module 20 classifies the data flows as: trusted data flows (…). Failed data stream () ), and security alert streams ( ).

[0041] Step S40: Dual-loop parallel processing. PCV gateway module 20 will process the trusted data stream ( The route is routed to the real-time energy efficiency control submodule 31 in the dual-loop control execution module 30.

[0042] PCV gateway module 20 will fail the data stream ( The route is routed to the long-term evolution and constraint generation submodule 32 in the dual-loop control execution module 30.

[0043] PCV gateway module 20 will transmit security alarm streams ( Routing to security response module 40.

[0044] The real-time energy efficiency control submodule 31 performs real-time operating condition adjustment of the GSHP system based on the received trusted data stream, and (in one embodiment) transmits the trusted data stream ( The message is forwarded to the long-term evolution and constraint generation submodule 32. The security response module 40 executes a security degradation strategy based on the received security alarm stream.

[0045] Step S50: Evolution and Feedback. The long-term evolution and constraint generation submodule 32 performs long-term window cumulative deviation analysis on trusted data streams (e.g., forwarded from the real-time energy efficiency control submodule 31).

[0046] When this cumulative deviation meets a specific evolutionary condition (e.g., exceeding a set constraint evolutionary threshold), The long-term evolution and constraint generation submodule 32 generates one or more dynamic physical constraints (DPCs).

[0047] The generated dynamic physical constraints are fed back to the PCV gateway module 20 to update its physical constraint set (R), thereby enabling the physical constraint verification logic in step S20 to adapt to the long-term physical characteristic changes of the GSHP system, thus forming a closed loop of the method.

[0048] In this embodiment of the invention, the specific implementation of the trusted data acquisition module 10 involves hardware-level identity binding and key generation for the sensor. This process ensures that each data acquisition source (sensor) in the GSHP system has a unique, tamper-proof identity identifier based on hardware physical characteristics.

[0049] Physically non-cloning function units are integrated into the hardware circuitry of sensors or data transmission units (DTUs) in ground source heat pump systems. These units utilize uncontrollable microscopic physical differences introduced during semiconductor manufacturing (e.g., random mismatches in transistor threshold voltages, subtle variations in gate delay) to produce unique physical characteristics.

[0050] In one specific embodiment, the physically unclonable function unit (PUF) can be a static random access memory (PUF). The PUF utilizes the random but chip-specific (biased to 0 or 1) initial state of an SRAM cell (composed of cross-coupled inverters) upon power-up as a physical fingerprint of the device. In another embodiment, an arbitrator PUF can also be used, which utilizes the result of a delay contention between two parallel paths as a response.

[0051] In a preferred embodiment, the key generation process is completed during the sensor manufacturing or system initialization (i.e., registration) phase. This process may include the following steps:

[0052] (1) Apply one or a set of defined excitation vectors Given physically non-clonable function units. For memory PUFs, the excitation vector... This refers to powering on or reading from the SRAM array; for the arbitrator PUF, the activation vector... It is a set of selection bits used to select the delay path.

[0053] ,in, It is an activation vector. It is the first An incentive.

[0054] (2) Physically non-clonable function element for excitation vector Generate a corresponding, unique, but potentially noisy, original physical fingerprint response. .

[0055] ,in It is the original physical fingerprint response of a physically non-clonable function unit. The excitation response function represents a physically non-clonable function unit.

[0056] (3) Response to the original physical fingerprint Processing is performed to generate a stable asymmetric key pair (private key). and public key Due to the original physical fingerprint response Readings may exhibit slight differences (noise) under different environmental conditions (such as temperature), and this processing step preferably includes a fuzzy extractor. The process is as follows:

[0057] ,in, It is the registration algorithm for the fuzzy extractor, which extracts data from the original physical fingerprint response. Extract a stable, high-entropy key seed. And at the same time generate public auxiliary data. Auxiliary data Do not disclose the key seed Or the original physical fingerprint response Any information, but used to correct the original physical fingerprint response in subsequent operations. The noise.

[0058] (4) Use this stable key seed As input, a deterministic key derivation function (KDF) or an asymmetric key generation algorithm is applied. Generate the final asymmetric key pair:

[0059] ,in, It is a key generation function. It is the sensor's hardware private key. That is the corresponding public key.

[0060] Generated private key It is securely stored inside the tamper-proof hardware of the sensor or DTU (e.g., a Secure Element (SE) or a Trusted Platform Module (TPM)) and is configured not to be read from the outside.

[0061] The generated public key Along with sensor labeling It is extracted. In some embodiments, auxiliary data It was also extracted along with the public key. and sensor identification Securely pre-registered into the public key store of PCV gateway module 20 for subsequent signature verification. Auxiliary data. It can be stored in local non-volatile memory, or it can also be transmitted to the PCV gateway module 20 for storage, for subsequent response and error correction.

[0062] For the specific circuit implementation of physically unclonable function units, fuzzy extractors Specific algorithms (such as error-correcting code-based algorithms) and asymmetric key generation algorithms (For example, algorithms based on ECC elliptic curve cryptography) Those skilled in the art can use existing mature solutions, the specific implementation of which is a well-known technology in the field and will not be described in detail here.

[0063] In this embodiment of the invention, after completing hardware-level identity binding, the trusted data acquisition module 10 performs trusted data packet generation and digital signature.

[0064] The trusted data acquisition module 10 periodically or when triggered collects raw operating data of the ground source heat pump system. For example, temperature, pressure, or flow rate values.

[0065] The trusted data acquisition module 10 immediately acquires the raw operating data. Add a timestamp and the sensor's identifier To build a data packet locally (e.g., in the processor of a sensor or DTU). .

[0066] ,in It is the original running data. It's a timestamp. It is the sensor's identifier.

[0067] In generating data packets Then, the trusted data acquisition module 10 immediately encrypts it to generate a digital signature. In a preferred embodiment, the process employs a method of signing the hash value of the data to improve computational efficiency.

[0068] (1) The trusted data acquisition module 10 applies a hash function. (e.g., SHA-256) Calculate data packets hash value .

[0069] ,in, It is a hash value. It is a hash function.

[0070] (2) The trusted data acquisition module 10 then calls its internally stored, unreadable private key. (Its generation method is as described above), and it uses a digital signature algorithm. (e.g., ECDSA) this hash value Perform a signature and generate a digital signature. .

[0071] ,in, It's a digital signature. It is a digital signature algorithm. It is the private key.

[0072] The trusted data acquisition module 10 will ultimately process the original data packets. Its corresponding digital signature Combined, they form a complete data transmission unit. And send it to PCV gateway module 20.

[0073] For hash functions and digital signature algorithms For specific implementation, those skilled in the art can use existing cryptographic standards, and their specific implementation is a well-known technology in this field, so it will not be described in detail here.

[0074] In this embodiment of the invention, the PCV gateway module 20 receives a data transmission unit from the trusted data acquisition module 10. Then, the first layer of verification in the dual security and physical verification logic is triggered, namely the security constraint verification. This verification aims to confirm the authenticity of the data packet's origin, the integrity of its content, and its timeliness.

[0075] PCV gateway module 20 parses the received data packets. Extract the sensor identifier from it. and timestamp .

[0076] Before performing signature verification, the PCV gateway module 20 preferably performs a replay attack check. The PCV gateway module 20 obtains the current system time. And calculate its relationship with the data packet timestamp. The difference. If the difference exceeds the preset reasonable time window threshold. If the data packet fails to pass the security constraint verification (e.g., 5 seconds), it is directly determined that the data packet has failed the security constraint verification and is marked as "expired data".

[0077] ;

[0078] If the timestamp check passes, the PCV gateway module 20 will then use the extracted sensor identifier. Retrieve the corresponding pre-registered public key from its securely stored public key store. .

[0079] PCV gateway module 20 uses the same hash function as trusted data acquisition module 10. For the received data packets Recalculate the hash value, that is .

[0080] ;

[0081] PCV gateway module 20 calls the standard signature verification algorithm. Using the retrieved public key Recalculated hash value and the received digital signature Perform verification to obtain the security constraint verification results. .

[0082] ,in, It is a Boolean value. If If it is True, it indicates that the data packet If the data is indeed generated by a legitimate sensor holding the corresponding private key and has not been tampered with during transmission, then the security constraint verification has been passed; if If the result is false, it indicates that the data source is unreliable or the content has been tampered with.

[0083] Only when Only when true will the PCV gateway module 20 process the data packet. The data packet is then passed to the next level for physical constraint verification. Any data packet that fails security constraint verification (including timestamp check failure or signature verification failure) will be directly marked as a "security alert flow" and will not be subject to further physical reasonableness judgment.

[0084] For signature verification algorithms The specific implementation corresponds to the aforementioned signature generation algorithm (such as ECDSA verification), which is a well-known technology in this field and will not be elaborated here.

[0085] For those who have passed the aforementioned security constraint verification (i.e.) (True) data packets The PCV gateway module 20 immediately initiates the second layer of verification, namely physical constraint (content rationality) verification.

[0086] PCV gateway module 20 extracts the data packet. Raw running data The PCV gateway module 20 maintains a set of physical constraints ( ), the physical constraint set ( It contains a set of rules for judging the physical validity of data.

[0087] In one embodiment, the physical constraint set ( This includes static threshold constraints. For example, the outlet temperature of buried pipes. raw running data It must be within a reasonable range based on the laws of physics. .like If the range is exceeded, the constraint is violated.

[0088] In another embodiment, the physical constraint set ( This includes multivariate correlation constraints. For example, for a ground source heat pump unit, the inlet water temperature on its evaporator side... and outlet water temperature raw running data The first law of thermodynamics must be satisfied, that is, under refrigeration conditions. If the original running data show If so, then the constraint is violated.

[0089] In yet another embodiment, the physical constraint set ( This includes timing logic constraints. For example, the raw operating data of a sensor at the current moment. (e.g., instantaneous flow rate) and the original operating data from the previous moment. The rate of change between them must be less than or equal to a preset maximum rate of physical change. .

[0090] ,in This is the original running data at the current moment. It is the original running data from the previous moment. It is the sampling time interval. This is the maximum rate of physical change. If the inequality does not hold, it is determined to be a violation of the constraint.

[0091] Physical constraint set ( It is dynamically updated. In addition to the static constraints set based on prior physical knowledge (such as the laws of thermodynamics) mentioned above, it also includes dynamic physical constraints (DPC) generated by the long-term evolution and constraint generation submodule 32 (as described below).

[0092] PCV gateway module 20 will display the original operating data As input, the physical constraint set is applied ( One or more relevant constraint rules in ) Perform verification to obtain physical verification results. .

[0093] ,in, Representative application physical constraint set ( The verification function. If If all relevant constraints are satisfied, then the physical verification result is... If any constraint is violated, the physical verification result is true; otherwise, the result is false. It is false.

[0094] This physical verification result This will be used for subsequent heterogeneous data flow routing. If If true, the data packet Classified as trusted data stream ( ).like The data packet is false (i.e., the data content does not conform to physical plausibility). Classified as failed data stream ( ).

[0095] As mentioned before, the set of physical constraints used by PCV gateway module 20 ( In one embodiment of the invention, it is divided into a static physical constraint set ( ) and a dynamic physical constraint set ( ).

[0096] Static physical constraint set ( ) is a set of physical constraints ( A subset of ), which is built during the initial deployment or initialization phase of the system and preloaded into the PCV gateway module 20.

[0097] Static physical constraint set ( The constraints contained herein are defined based on prior knowledge of the inherent physical characteristics, equipment specifications, and system design logic of the ground source heat pump system that do not change significantly over time.

[0098] In a specific embodiment, the static physical constraint set ( This includes "device range constraints." These constraints are based on the technical specifications provided by the sensor or actuator manufacturer. For example, a temperature sensor deployed in a buried pipe loop. Its original operating data It must be within its physical range Within.

[0099] ,in, This is the lower limit of the sensor's specifications (e.g., -10℃). This is the upper limit of the sensor's specifications (e.g., +50°C). Any reading outside this range violates this static constraint.

[0100] In another embodiment, the static physical constraint set ( This includes "thermodynamic logic constraints." These constraints are based on the laws of thermodynamics. For example, for a ground source heat pump unit operating in cooling mode, its evaporator (user side)'s raw operating data... Must meet:

[0101] ,in, It is the inlet water temperature of the evaporator. This is the outlet water temperature of the evaporator. If If this violates the basic physical laws of heat exchange, it is considered invalid data.

[0102] In yet another embodiment, the static physical constraint set ( This includes "fluid conservation constraints." These constraints are based on the law of mass conservation in fluid dynamics. For example, for a tee junction in a system, its original operating data... Must meet:

[0103] ,in and It is the inflow of traffic from the two branch roads. This is the main outgoing traffic of this node. This is a very small flow balance tolerance set to account for sensor measurement errors. If this equation does not hold, then the law of conservation of mass is violated.

[0104] Static physical constraint set ( As the initial and fundamental basis for judging the physical rationality of the system, it is used to perform the physical constraint verification in the aforementioned steps.

[0105] In this embodiment of the invention, the PCV gateway module 20 verifies the security constraints generated in the aforementioned steps. and physical verification results For the received data packets Perform classification and routing of heterogeneous data streams.

[0106] This routing mechanism is based on a defined set of logical rules.

[0107] (1) If the safety constraint verification result If the packet is false (as described above due to signature verification failure or timestamp expiration), the PCV gateway module 20 will send the packet to the appropriate location. Immediately categorize as a security alarm flow ( ).

[0108] The security alert stream ( The signal is routed to the security response module 40 to trigger a security alarm or a system degradation strategy.

[0109] (2) If the safety constraint verification result It is true, and the physical verification result is true. If the data is True (as described above, it satisfies both security and physical plausibility), the PCV gateway module 20 will send the data packet... Classified as trusted data stream ( ).

[0110] This trusted data stream ( It is routed to the real-time energy efficiency control submodule 31 in the dual-loop control execution module 30 and used as a valid input for the real-time energy efficiency control of the GSHP system.

[0111] (3) If the safety constraint verification result It is true, but the physical verification result is incorrect. If the data packet is false (as described above, the data source is credible but the content violates physical constraints), the PCV gateway module 20 will reject the data packet. Classified as failed data stream ( ).

[0112] The failed data stream ( The data stream is routed to the long-term evolution and constraint generation submodule 32 within the dual-loop control execution module 30. This data stream is not used for real-time control, but rather as a log and diagnostic basis for physical faults (e.g., sensor momentary failure or communication anomalies). Analysis and constraint evolution for analyzing system physical characteristics (such as sensor drift or pipe scaling) are instead based on a trusted data stream (forwarded by the real-time energy efficiency control submodule 31). )conduct.

[0113] In this embodiment of the invention, the real-time energy efficiency control submodule 31 included in the dual-loop control execution module 30 is responsible for executing the first-level control loop, namely real-time energy efficiency control.

[0114] Real-time energy efficiency control submodule 31 specifically receives and only receives trusted data streams routed from PCV gateway module 20. ).

[0115] Real-time energy efficiency control submodule 31 receives data from this trusted data stream ( Parse the data packet from ) And extract the original runtime data that has been double-verified. This is the original runtime data. (For example, the current inlet and outlet water temperature, flow rate, pressure, etc.) are used as reliable inputs for the real-time energy efficiency control algorithm.

[0116] The goal of this real-time energy efficiency control algorithm is to minimize the total power consumption of the GSHP system by optimizing system operating parameters, while meeting the current load demand of the GSHP system. The real-time energy efficiency control submodule 31 runs the algorithm to calculate a set of optimal control commands.

[0117] In one specific embodiment, the control algorithm is constructed as a real-time optimization (RTO) problem. Its objective function is... It is to minimize the total power consumption of the system. ,Should It is based on the original running data It is calculated using the system energy consumption model.

[0118] ,in, The control commands (e.g., the frequency of the ground source pump, the frequency of the user pump, and the frequency of the compressor) serve as the variables to be optimized in this optimization problem. It is the raw runtime data that serves as input; It is about minimizing the total power consumption of the system.

[0119] The optimization process is subject to a set of constraints, which are also based on the original operating data. Conduct an assessment:

[0120] (1) Load meets constraints: For example, the main water supply temperature sensor on the user side (A piece of raw operational data) must be maintained at the demand setpoint. One allowable deviation Within the range, that is .

[0121] (2) Equipment safety constraints: For example, control commands as variables to be optimized. Various parameters (such as frequency) It must be within the physical operating range allowed by the equipment. Inside.

[0122] The real-time energy efficiency control submodule 31 solves this optimization problem to obtain the optimal control command. .

[0123] ,in Is this optimization problem based on the current original running data? The solution below.

[0124] The optimal control command It is then sent to the corresponding system actuator 50 of the GSHP system (e.g., the frequency converter of the water pump or the driver of the compressor) to adjust the actual operating state of the GSHP system.

[0125] For the specific construction of real-time optimization algorithms (such as Sequential Quadratic Programming, SQP) or energy consumption models, those skilled in the art can use existing mature solutions, the specific implementation of which is well-known in the field and will not be elaborated here.

[0126] In this embodiment of the invention, the dual-loop control execution module 30 includes a long-term evolution and constraint generation submodule 32, which is responsible for executing the second-level control loop, namely the long-term evolution loop. The purpose of this loop is to analyze the changes in system characteristics caused by physical aging (such as pipe scaling, heat exchanger efficiency reduction) or sensor drift.

[0127] The first step in this long-term evolutionary cycle is to extract the model's measured bias vector. This vector is used to quantify the difference between the system's current actual physical behavior and its theoretical model behavior.

[0128] Long-term evolution and constraint generation submodule 32 continuously receives and buffers trusted data streams forwarded by real-time energy efficiency control submodule 31. (Includes raw runtime data) and optimal control commands .

[0129] (As described above) This long-term evolution and constraint generation submodule 32 can also receive failed data streams routed from the PCV gateway module 20. And it is used for logging and diagnostic analysis of physical faults.

[0130] The long-term evolution and constraint generation submodule 32 internally maintains a GSHP physics model. The GSHP physical model It is a mechanistic model or high-precision proxy model that describes how the system state (such as outlet temperature) responds to changes in input state (such as inlet temperature, flow rate) and control commands (such as pump frequency) in an “ideal” or “initial” state (i.e., without aging or drift).

[0131] Model measured deviation vector The extraction process specifically includes:

[0132] (S1) The long-term evolution and constraint generation submodule 32 collects data from... Real-time, originating from trusted data streams ( ) system input state variables (For example, the inlet water temperature and flow rate of the heat exchanger) and the data generated by the real-time energy efficiency control submodule 31. Optimal control command at any time (For example, water pump frequency).

[0133] (S2) This long-term evolution and constraint generation submodule 32 will and As input, it is substituted into its internal GSHP physics model. ,calculate Model prediction at time .

[0134] ,in yes The model prediction at time (a vector containing multiple predictor variables, for example) ); It is the GSHP physics model; yes The system input state variables at any given time; yes The optimal control command at any given time.

[0135] (S3) This long-term evolution and constraint generation submodule 32 from Reliable data stream at any given time ( In the process, extract the values ​​predicted by the model. Corresponding measured state values (For example, (Measured water temperature at any given time).

[0136] (S4) The long-term evolution and constraint generation submodule 32 calculates the measured state values ​​point by point. and model predictions Deviation between .

[0137] ,in It is the first The deviation values ​​of each state variable; It is the measured state value of the variable; This is the model prediction value for that variable.

[0138] (S5) This long-term evolution and constraint generation submodule 32 will include all Deviation values ​​of key state variables Combine, form Model-measured deviation vector at time step .

[0139] ,in That is The measured deviation vector of the model at time step.

[0140] The measured deviation vector of this model It is continuously generated and stored for subsequent system aging state assessment and physical constraint evolution.

[0141] For the GSHP physical model The specific construction (e.g., a mechanistic model based on heat transfer and fluid dynamics or a data-driven proxy model) can be achieved by those skilled in the art using existing mature solutions, and their specific implementation is well-known in the field, so it will not be elaborated here.

[0142] In this embodiment of the invention, the long-term evolution and constraint generation submodule 32 obtained Model-measured deviation vector at time step Subsequently, the long-term evolution and constraint generation submodule 32 does not immediately respond to instantaneous deviations, but instead performs a cumulative analysis of the deviations. This step aims to distinguish between sporadic measurement noise and persistent deviation trends that indicate changes in the physical properties of the system.

[0143] The long-term evolution and constraint generation submodule 32 collects and buffers data within a preset time window. The sequence of model measured bias vectors generated within (e.g., over the past 24 hours) .

[0144] The long-term evolution and constraint generation submodule 32 applies a statistical function. Process the sequence of deviation vectors to compute a scalar cumulative deviation metric. This cumulative deviation measure Reflects on the time window The actual behavior of the internal system deviates from its initial physical model by an overall margin.

[0145] In one specific embodiment, the statistical function Calculate the time window Internal model measured deviation vector The moving average of the L2 norm (i.e., the Euclidean length of the vector):

[0146] ,in, yes The cumulative deviation measure over time; It is the size of the time window; yes The L2 norm of the measured deviation vector of the time-time model is calculated as follows:

[0147] ,in yes The first time deviation vector Each component.

[0148] In another embodiment, to improve the smoothness of computation and reduce storage requirements, this cumulative deviation metric... It can also be calculated using the Exponential Moving Average (EMA):

[0149] ,in It is a smoothing factor between 0 and 1.

[0150] The long-term evolution and constraint generation submodule 32 internally presets two thresholds for judgment: a model tuning threshold. and a constraint evolution threshold These two thresholds define the tolerance range for systematic bias and satisfy the following conditions: .

[0151] The long-term evolution and constraint generation submodule 32 continuously calculates the cumulative deviation metric. Compare with these two thresholds to trigger different response logic:

[0152] (1) If This indicates that the current model's measured bias vector is statistically within an acceptable noise range. The system is determined to be operating on the initial GSHP physics model. As expected, submodule 32 does not perform any action.

[0153] (2) If This indicates that the system has experienced a persistent and non-negligible deviation, but the magnitude of this deviation has not yet reached the point of fundamental change. This situation typically corresponds to a slight sensor drift or an initial, slow decline in heat exchanger efficiency. The long-term evolution and constraint generation submodule 32 determines that "model tuning" is needed and generates a model tuning trigger signal. .

[0154] (3) If This indicates that the system has experienced a significant and sustained deviation. This situation corresponds to a major change in the system's physical characteristics, such as severe scaling in the buried pipe system leading to a significant decrease in heat transfer performance, or the complete failure of critical sensors. The long-term evolution and constraint generation submodule 32 determines that "constraint evolution" is required and generates a constraint evolution trigger signal. .

[0155] These two trigger signals and This will initiate subsequent GSHP physics model parameter adjustments or physics constraint sets ( Iterative updates of ).

[0156] In this embodiment of the invention, when the long-term evolution and constraint generation submodule 32 determines the cumulative deviation metric... Meet the conditions When (as described above), a model tuning trigger signal will be generated. .

[0157] The model is tuned to trigger signals. This will launch a GSHP physics model. This is a parameter fine-tuning procedure. The purpose of this procedure is to adjust the GSHP physics model. An internal set of adjustable parameters , so that it ( The predicted behavior is re-aligned with the current actual physical behavior of the system (by trusted data flow). (Reflects) Matches.

[0158] GSHP physical model Internal adjustable parameters , refers to the key coefficients in the model used to characterize the physical properties of the system.

[0159] In one embodiment, The total heat transfer coefficient can include one or more heat exchangers. When cumulative deviation is measured Continue to exceed At this time, it may indicate that a slight fouling buildup has occurred in the heat exchanger, leading to its actual... Value compared to the initial design value A decline occurred.

[0160] In another embodiment, This can include a specific temperature sensor Zero drift coefficient or gain drift coefficient For example, the GSHP physical model Corrected reading of the sensor It can be represented as .when Exceed When the deviation is mainly contributed by this sensor, the system will fine-tune. or .

[0161] The parameter fine-tuning procedure is constructed as a parameter estimation or optimization problem. The long-term evolution and constraint generation submodule 32 utilizes time windows. Internally accumulated measured state values and the corresponding system input state variables and optimal control commands Data set, solve for a set of optimal model parameters .

[0162] The optimization problem aims to minimize the sum of squared residuals (RSS) between the model predictions and the observed values:

[0163] ,in These are the optimal model parameters obtained by solving the problem; These are adjustable parameters that need to be optimized; yes The measured state value at any given time; The prediction function of the GSHP physics model with parameters.

[0164] Long-term evolution and constraint generation submodule 32 calculates the optimal model parameters. Then, use renew Original internal parameters .

[0165] GSHP physical model After the update, the model predictions generated in the aforementioned steps will be... This will be closer to the actual measured value. This makes the cumulative deviation measure calculated in the preceding steps... Falling back to the model tuning threshold The following section demonstrates how the model adapts to slight changes in the system.

[0166] Meanwhile, the calculation of the objective function is as described in the preceding steps. (i.e., minimize the total power consumption of the system) The system energy consumption model is the GSHP physical model. Part or all. When Parameters (e.g., overall heat transfer coefficient) After being fine-tuned and updated, the real-time energy efficiency control submodule 31 will automatically call this updated model when performing the real-time optimization of the aforementioned steps.

[0167] This ensures that real-time energy efficiency control (first-level cycle) is based on a model that best reflects the current physical state of the system (e.g., slight scaling has occurred), thereby improving the optimal control command. The accuracy of the system and its actual operational efficiency. This process constitutes "control evolution".

[0168] For parameter estimation algorithms (such as recursive least squares, gradient descent, etc.), those skilled in the art can use existing mature solutions, the specific implementation of which is a well-known technology in the field, and will not be described in detail here.

[0169] In this embodiment of the invention, when the long-term evolution and constraint generation submodule 32 determines the cumulative deviation metric... Meet the conditions When (as described in the preceding steps), a constrained evolution trigger signal will be generated. .

[0170] This constraint triggers the evolution signal. The generation of this indicates that the GSHP system is no longer in a state of slight drift, but rather its core physical characteristics (e.g., the overall heat transfer coefficient of the buried pipe heat exchanger) are being utilized. or the efficiency of the water pump The material has undergone significant and irreversible changes due to long-term operation (such as scaling and wear).

[0171] In this case, the static physical constraint set as described in the preceding steps ( (Its construction based on the system's "initial" or "ideal" state) is no longer applicable. Continuing to use this static physical constraint set ( This will lead to a large number of genuine sources ( (True) and data packets reflecting the current aging state were incorrectly identified as invalid data streams. () (False).

[0172] Therefore, constrained evolution triggering signals A dynamic physical constraint (DPC) generation procedure will be initiated. The purpose of this procedure is to generate a new set of physical constraints, which are more lenient or have changed form, based on the system's current, aged operational data, to replace the original static constraints.

[0173] Long-term evolution and constraint generation submodule 32 accesses its time window Internal (i.e., trigger) The accumulated trusted data stream before the signal ( (Includes raw runtime data) The dataset.

[0174] In one embodiment, the program generates a new "temporal logic constraint" by statistically analyzing the dataset. This constraint replaces the original static temporal logic constraint based on the initial state. For example, for a variable that has experienced sensor drift, its current raw operating data... Maximum physical rate of change This may have changed. The long-term evolution and constraint generation submodule 32 recalculates the actual maximum rate of change of the dataset under this operating condition, and adds a safety margin. This generates a new Dynamic Physical Constraint (DPC). :

[0175] ,in .

[0176] In another embodiment, the procedure is based on the evolved GSHP physical model whose parameters have been fine-tuned in the aforementioned steps. This generates a new "multivariate correlation constraint" (corresponding to the description of the preceding steps). For example, for a scaled heat exchanger, the temperature difference between its inlet and outlet water... Compared to the initial state, under the same flow rate The value has decreased significantly.

[0177] Long-term evolution and constraint generation submodule 32 runs the evolved Calculate under different inputs (e.g. Under these conditions, the model prediction value corresponding to the current physical state of the system. A reasonable range (e.g., a confidence interval) ).

[0178] The long-term evolution and constraint generation submodule 32 encapsulates the new reasonable range obtained from this calculation into a dynamic physical constraint (DPC). :

[0179] ,in This is the data to be inspected. From the evolved model The exported constraint functions.

[0180] For Dynamic Physical Constraints (DPC) (such as , The specific generation algorithm (e.g., based on kernel density estimation, regression analysis or machine learning methods) can be obtained by those skilled in the art using existing mature solutions, and its specific implementation is a well-known technology in the field, and will not be described in detail here.

[0181] The core of this generation process lies in the feedback step: the long-term evolution and constraint generation submodule 32 will generate new dynamic physical constraints (DPCs) (such as...) , Send to PCV gateway module 20.

[0182] PCV gateway module 20 receives the Then, use it to update the physical constraint set ( In one embodiment, the The physical constraint set will be replaced. The existing static physical constraints (from) the same variable in the original ) ).

[0183] Subsequently, when performing the physical constraint verification in the aforementioned steps, the PCV gateway module 20 will use this evolved Dynamic Physical Constraint (DPC). As The basis for judgment.

[0184] This feedback mechanism ensures that the physical constraint verification (second-level verification) standard is adaptable to the long-term physical aging of the GSHP system. This allows the PCV gateway module 20 to accurately distinguish between "trustworthy data reflecting the aging state" and "actual failure data occurring in the aging state" even after the system has been running for, for example, five years (with performance degradation), thus completing the closed loop of the entire long-term evolution cycle.

[0185] In this embodiment of the invention, the security response module 40 is responsible for receiving and processing the security alarm stream routed from the PCV gateway module 20. ).

[0186] As described in the preceding steps, when a data packet Because the security constraint verification in the aforementioned steps failed (i.e. (Because it is fake) and thus classified as a security alert stream ( When this occurs, the data stream is immediately sent to the security response module 40.

[0187] The security response module 40 receives the data packet. Next, the data packet will be parsed to extract its data source identifier. ), and the resulting The specific security failure reason for being identified as fake (e.g., "digital signature verification failed" or "timestamp exceeded tolerance window").

[0188] The security response module 40 first performs the recording and reporting of security events. In one embodiment, the module 40 reports the security event (including...) The data (timestamp, failure reason, and data packet content) is stored in a security log database for subsequent security auditing. Simultaneously, the security response module 40 generates a security alarm, which is sent to the upper-layer Supervisory Control and Data Acquisition (SCADA) system via an industrial communication protocol (e.g., Modbus TCP or OPC UA) to notify maintenance personnel.

[0189] In one specific embodiment, the core function of the security response module 40 is to execute a data source isolation policy to prevent the spread of security threats.

[0190] The data source isolation strategy specifically includes:

[0191] (S1) Security Response Module 40 maintains a data source blacklist. Upon receiving a security alarm stream ( After that, the security response module 40 identifies the data source from which the data packet originated. Add to the blacklist of this data source ( )middle.

[0192] (S2) The security response module 40 will update the data source blacklist ( Feedback is sent to PCV gateway module 20.

[0193] (S3) The PCV gateway module 20 adds a blacklist check at the very beginning of its security constraint verification. This applies to any subsequently received data packets. The PCV gateway module 20 first checks its Does it exist in the data source blacklist ( )middle.

[0194] (S4) If the Existence In the process, the PCV gateway module 20 will immediately determine the security constraint verification result of the data packet. If the value is false, it is routed to the security alert stream. Instead of performing the signature verification or timestamp verification steps mentioned above, the aforementioned steps are no longer performed.

[0195] This isolation mechanism ensures that all subsequent data packets from data sources that have been identified as insecure (e.g., whose security keys have been compromised or whose clocks have been severely faulty) are immediately intercepted and cannot enter the trusted data stream described in the preceding steps. or failed data stream () This allows for the rapid blocking of security threats.

[0196] In another embodiment, the security response module 40 is also responsible for triggering the system's security degradation operation mode. If isolated... The absence of a key sensor in the corresponding GSHP system (e.g., the user-side main water supply temperature sensor) will prevent the Real-Time Optimization (RTO) in the aforementioned steps from being performed.

[0197] At this time, the safety response module 40 sends a degraded operation trigger signal to the real-time energy efficiency control submodule 31. Upon receiving this signal, the real-time energy efficiency control submodule 31 will suspend the execution based on... The model's real-time optimization algorithm instead uses a set of preset, fixed safety control parameters that do not depend on the failed sensor (e.g., locking the ground source pump and user-side pump at a fixed frequency of 50% and using simple start-stop logic to control the compressor) to ensure the continuous operation of the GSHP system's basic cooling or heating functions under safety threats.

[0198] In this embodiment of the invention, the security response module 40 performs a judgment, that is, confirms the data source that has been blocked by the aforementioned data source isolation strategy ( When a data source is a "critical data source" necessary for the operation of the GSHP system, a security degradation of the control policy will be triggered.

[0199] "Key data source" refers to the raw operational data it provides. This is the input required for the real-time energy efficiency control submodule 31 to execute the aforementioned real-time optimization (RTO) algorithm. For example, this data source is the main water supply temperature sensor on which the aforementioned "load satisfaction constraint" depends. .

[0200] In this situation, the safety response module 40 sends a degraded operation trigger signal to the real-time energy efficiency control submodule 31.

[0201] Upon receiving the degradation operation trigger signal, the real-time energy efficiency control submodule 31 immediately executes the safety degradation of the control strategy. This degradation process specifically includes:

[0202] (S1) The real-time energy efficiency control submodule 31 immediately suspends or terminates the currently running real-time optimization (RTO) algorithm. This operation is necessary because the algorithm's input (from the "critical data source") is no longer reliable or has been interrupted; continuing to execute the optimization algorithm would lead to the calculation of incorrect or even dangerous optimal control commands. .

[0203] (S2) The real-time energy efficiency control submodule 31 loads a predefined safety degradation control strategy from its internal storage.

[0204] This security degradation control strategy is a robust and simplified control logic that does not rely on isolated "critical data sources." The goal of this strategy is no longer to minimize the total power consumption of the system, as pursued in the aforementioned real-time optimization. (Energy efficiency) is not the goal, but rather to ensure the basic functions of the GSHP system (such as cooling / heating) and equipment safety.

[0205] (S3) The real-time energy efficiency control submodule 31 generates a set of fixed safety control parameters based on the safety degradation control strategy. and the This is sent as a control command to the system actuator 50.

[0206] In one specific embodiment, the security degradation control strategy may include the following rules:

[0207] (1) Set the frequency of the ground source pump and the frequency of the user pump, which were originally "variables to be optimized" in the aforementioned real-time optimization, to fixed, load-independent safety values, for example... That is, all of them are locked at 50% frequency.

[0208] (2) The compressor's control logic is shifted from relying on "critical data sources" (such as...) The optimized control of the system is switched to simple on-off control based on a non-critical, reliable auxiliary sensor (e.g., a backup return water temperature sensor).

[0209] This security degradation mechanism ensures that the GSHP system is protected against security attacks targeting critical sensors (such as data tampering or replay attacks, which could lead to security constraint verification). When the condition is false, the control system (dual-cycle control execution module 30) can automatically switch from the "optimal energy efficiency" mode to the "safety priority" mode, thereby maintaining the basic operation of the system while ensuring equipment safety.

Claims

1. A method for intelligent monitoring and energy efficiency management of the entire process of a ground source heat pump system, characterized in that, Includes the following steps: The system collects raw operating data of the ground source heat pump system and uses a physically unclonable function unit integrated into the data acquisition terminal to generate a hardware-level digital signature, which constitutes a data packet carrying the signature. Perform dual verification on the data packet carrying the signature, the dual verification including: Security constraint verification based on a pre-stored public key to verify the hardware-level digital signature, and physical constraint verification by comparing the original running data with the physical constraint set after the security constraint verification is passed. Based on the results of the dual verification, the data streams are classified into trusted data streams, failed data streams, and safety alarm streams. The trusted data streams are used for real-time operating condition adjustment and energy efficiency control of the ground source heat pump system, forming the first control loop. The trusted data streams and failed data streams are used for long-term window cumulative analysis. The failed data streams are used for physical fault diagnosis and log recording, and the trusted data streams are used to analyze long-term changes in the system's physical characteristics, forming the second control loop. In the second control loop, when the cumulative analysis results of the model measured deviations contained in the trusted data streams meet the preset constraint evolution conditions, dynamic physical constraints are generated, and the physical constraint set is updated using the feedback of the dynamic physical constraints.

2. The method for intelligent monitoring and energy efficiency management of a ground source heat pump system according to claim 1, characterized in that, The step of generating a hardware-level digital signature using a physically unclonable function unit specifically includes: applying a hash function to calculate the hash value of the data packet consisting of the original running data, timestamp, and sensor identifier; and calling the private key generated by the physically unclonable function unit to sign the hash value, thereby generating the hardware-level digital signature.

3. The method for intelligent monitoring and energy efficiency management of a ground source heat pump system according to claim 1, characterized in that, Before performing signature verification, the security constraint verification also includes: obtaining the current system time and determining whether the difference between the timestamp in the data packet and the current system time exceeds a preset time window threshold. If it exceeds the threshold, the security constraint verification is deemed to have failed.

4. The method for intelligent monitoring and energy efficiency management of a ground source heat pump system according to claim 1, characterized in that, The set of physical constraints includes at least one of the following constraints: static threshold constraints based on the physical range of the device; multivariable correlation constraints based on the laws of thermodynamics; or temporal logic constraints based on the maximum physical rate of change.

5. The method for intelligent monitoring and energy efficiency management of a ground source heat pump system according to claim 1, characterized in that, The specific steps for classifying data streams into trusted data streams, failed data streams, and security alarm streams are as follows: if the security constraint verification fails, it is classified as a security alarm stream; if the security constraint verification passes and the physical constraint verification passes, it is classified as a trusted data stream. If the security constraint verification passes but the physical constraint verification fails, it is classified as a failed data stream.

6. The method for intelligent monitoring and energy efficiency management of a ground source heat pump system according to claim 1, characterized in that, The steps of real-time operating condition adjustment and energy efficiency control specifically include: taking the original operating data in the trusted data stream as input, solving a real-time optimization problem that minimizes the total power consumption of the system as the objective function and satisfies load demand constraints and equipment safety constraints, and generating optimal control commands.

7. The method for intelligent monitoring and energy efficiency management of a ground source heat pump system according to claim 1, characterized in that, The analysis of the measured deviation of the model specifically includes: substituting the system input state variables and optimal control commands from the trusted data stream into the preset ground source heat pump physical model to calculate the model prediction value; and calculating the deviation between the model prediction value and the measured state value in the trusted data stream to form a model measured deviation vector.

8. The method for intelligent monitoring and energy efficiency management of a ground source heat pump system according to claim 7, characterized in that, Specifically, satisfying the preset constraint evolution condition involves: performing cumulative analysis on the measured deviation vector of the model within a preset time window to obtain a cumulative deviation metric; and determining that the constraint evolution condition is satisfied when the cumulative deviation metric is greater than a preset constraint evolution threshold.

9. The method for intelligent monitoring and energy efficiency management of a ground source heat pump system according to claim 1, characterized in that, The step of updating the physical constraint set using the dynamic physical constraint feedback specifically includes: sending the newly generated dynamic physical constraint to the PCV gateway module that performs dual verification, and replacing the original static physical constraints for the same variable in the physical constraint set with the dynamic physical constraint.

10. The method for intelligent monitoring and energy efficiency management of a ground source heat pump system according to claim 1, characterized in that, The method further includes: routing the security alarm flow to the security response module, whereby the security response module adds the data source identifier of the data flow to the data source blacklist and feeds it back to the PCV gateway module that performs dual authentication, thereby intercepting all subsequent data packets from the data source.

Images