Water conservancy information security protection method and system

Abstract

The application provides a water conservancy information security protection method and system, comprising the following steps: S1 edge device physical security reinforcement, S2 dynamic encryption data collection, S3 adaptive fragmentation redundant transmission, S4 blockchain-distributed joint storage, S5 privacy secure data sharing, S6 multi-dimensional threat real-time disposal, S7 dynamic permission intelligent regulation and control; the application deploys a protection module for water conservancy sensors and RTU, monitors equipment vibration, uncovers and environmental parameters, detects abnormal physical attacks or environmental over-limit, automatically erases the key, starts the data protection mode, prevents data leakage when the equipment is physically damaged, ensures the data security of the edge device, and adopts MPC or federal learning technology for cross-system data interaction, the original data remains local, generates a blockchain audit log record to call the behavior, realizes data sharing while ensuring that the data does not leave the local, ensures data privacy, and provides data call traceability basis.

Description

Technical Field

[0001] This invention belongs to the field of information protection, specifically a water conservancy information security protection method and system. Background Technology

[0002] With the development of information technology, data security protection is of paramount importance. Water resources information differs significantly from general enterprise information. Water resources information involves numerous edge devices, such as water sensors and RTU remote terminal units, distributed in complex field environments, facing the risk of physical attacks such as floods and human sabotage. Furthermore, it has extremely high requirements for data real-time performance and integrity; data leaks or equipment malfunctions could trigger major security incidents. In contrast, general enterprise information is primarily stored in a relatively secure internal network environment, where security threats mainly come from network attacks and unauthorized internal access. Therefore, the focus of security protection for general enterprise information differs from that for water resources information.

[0003] For example, the information security protection method and system disclosed in publication number CN117852015B has shortcomings in access control within the field of information security protection. Regarding enterprise data access, one approach is to verify only device permissions. However, when a high-privilege user uses a low-privilege device to access high-privilege data, a mismatch occurs between the accessed data scope and the user's permissions. Another approach is to verify only user permissions. However, when a high-privilege user uses a low-privilege device to access low-privilege data, this adds an additional user authentication step, impacting the user experience.

[0004] In the context of water conservancy information, relying solely on device access verification cannot effectively protect data security when sensing devices are subjected to physical attacks or environmental limitations, potentially leading to key leaks and unauthorized data access. Conversely, relying solely on user access verification fails to provide timely data protection when devices face physical risks, and the additional authentication procedures can delay emergency response, failing to guarantee the security and integrity of water conservancy data under extreme conditions. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a water conservancy information security protection method and system, which solves the problems mentioned in the background art.

[0006] A water resources information security protection method includes the following steps:

[0007] S1 edge device physical security hardening: Deploy physical security protection modules in hydraulic sensors and RTU remote terminal units to monitor the vibration, open cover status and environmental parameters of the equipment in real time. When abnormal physical attacks or environmental limits are detected, the key is automatically erased and the data protection mode is activated.

[0008] S2 Dynamic Encrypted Data Acquisition: Based on the PUF Physical Unclonable Function, a unique fingerprint of the device is generated. Combined with environmental noise, an encryption key is dynamically generated. The collected water conservancy monitoring data is encrypted after adding a triplet identifier of device fingerprint, timestamp and geographic coordinates. The key validity period is set to 3-10 minutes.

[0009] S3 Adaptive Fragmented Redundant Transmission: The number of data fragments N≥3 is dynamically determined based on network latency and packet loss rate. The main transmission path is encrypted using the national cryptographic SM4 algorithm, while the backup path uses a lightweight encryption algorithm. The receiving end verifies data integrity through hash verification and blockchain evidence storage.

[0010] S4 Blockchain - Distributed Consortium Storage: Data hash values ​​are written to consortium blockchain nodes, the original data is encrypted and stored in fragments on the IPFS interplanetary file system, and overdue data is destroyed through smart contracts;

[0011] S5 privacy and security data sharing: When interacting with data across systems, it uses MPC secure multi-party computation or federated learning technology to ensure that the original data does not leave the local storage node and generates a blockchain audit log to record data call behavior;

[0012] S6 Multi-Dimensional Threat Real-Time Handling: Based on the user's historical operation sequence, a behavior baseline model is trained to detect abnormal access behavior in real time. When it is identified that an overseas IP is logging in outside of working hours or scanning high-risk ports, it will link with network security devices to generate an Access Control List (ACL) for millisecond-level blocking.

[0013] S7 Dynamic Permission Intelligent Control: Establishes a risk assessment matrix, calculates dynamic permission levels by comprehensively considering user role permissions, equipment health indicators, and network environment parameters. When the network latency of the reservoir monitoring point is greater than 200ms or the water level exceeds the warning line, the remote control command permission is automatically downgraded to read-only mode.

[0014] Preferably, the criteria for determining the abnormal physical attack include a vibration frequency ≥ 50Hz, a corresponding acceleration ≥ 5g, or the device casing being open for more than 2 seconds. The environmental parameters exceeding the limits include a temperature exceeding the range of -20℃ to +60℃, humidity > 90%RH, or GPS location offset ≥ 5 meters.

[0015] Preferably, the generation rule for the dynamic encryption key is that the key validity period is dynamically adjusted according to the water conservancy monitoring parameters. The greater the difference between the water level and the warning water level, the shorter the key validity period. When the dam displacement rate reaches 2 mm per hour or more, the key validity period will be shortened to 50% of the normal state. The encrypted data packet header includes water conservancy project metadata, including the monitoring point elevation, data acquisition time, and equipment installation coordinates. The key validity period T is dynamically correlated with the water conservancy monitoring parameters, satisfying the formula:

[0016] Preferably, the number of fragments is dynamically determined based on network quality; the higher the network latency, the more fragments are required. When the network latency exceeds 300 milliseconds, BeiDou short message service is forcibly used as a backup transmission path. The primary path uses SM4-GCM encryption mode, and the backup path uses the ChaCha20-Poly1305 algorithm. Keys are generated independently and are not reused. The number of fragments N is dynamically calculated based on network quality, satisfying the formula:

[0017]

[0018] Preferably, the user risk level of the risk assessment matrix is ​​divided into 1-3 levels, with level 1 being a regular queryer and level 3 being a system administrator. The equipment health is calculated based on the equipment vibration speed; the higher the vibration speed, the lower the equipment health. When the vibration speed exceeds a certain threshold, the equipment health is set to zero. The network environment risk coefficient is set to 1 when the network latency exceeds 200 milliseconds, and 0 otherwise. When the risk assessment result reaches 6 or above, high-risk operations such as adjusting the opening of the floodgate and starting / stopping the pumping station unit are prohibited. The calculation formula for the risk assessment matrix is: Risk = 0.5Puser + 0.3Sdevice + 0.2Nenu, where Puser is the user risk level, Sdevice is the equipment health, and Nenu is the network environment risk coefficient.

[0019] Preferably, when the gate motor temperature rises to 40K relative to the ambient temperature, or when the power supply voltage fluctuates by more than ±20% from the rated 24VDC, the equipment health is directly reset to zero. When the reservoir water level exceeds the design flood level, the network environment risk coefficient will be forcibly set to the maximum value of 1.

[0020] A water conservancy information security protection system includes the following modules:

[0021] Physical safety sensing module: Deployed in water conservancy terminal equipment, integrating anti-tamper self-destruct unit and environmental sensors to monitor the physical status of equipment in real time;

[0022] Dynamic encryption acquisition module: Generates a unique fingerprint for the device through the PUF (Physically Unclonable Function) and dynamically generates an encryption key based on environmental noise to encrypt the acquired water conservancy project data;

[0023] Adaptive fragmentation transmission module: Based on network latency, packet loss rate and water conservancy business priority, encrypted data is fragmented and transmitted through multiple paths, supporting dynamic switching between national cryptographic SM4 and ChaCha20 algorithms;

[0024] Distributed storage and evidence preservation module: It consists of blockchain nodes and IPFS storage cluster. The blockchain records data hash values ​​and operation logs, and IPFS storage is encrypted and sharded and executes smart contract-driven lifecycle management.

[0025] Privacy-preserving computation sharing module: Based on the MPC secure multi-party computation and federated learning framework, it enables joint analysis of water conservancy data across regions, with the raw data remaining in the local storage node;

[0026] Intelligent access control module: It has a built-in risk assessment matrix and user behavior baseline model, dynamically adjusts user operation permissions, and works with the firewall to execute real-time blocking policies.

[0027] Preferably, the water conservancy terminal equipment includes a water level gauge, a displacement sensor, and a gate controller, and the flood control command in the water conservancy operation is the highest level.

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] 1. This invention deploys a protection module in the hydraulic sensor and RTU to monitor equipment vibration, cover opening and environmental parameters. If abnormal physical attacks or environmental limits are detected, the key is automatically erased and the data protection mode is activated to prevent data leakage when the equipment is physically damaged, thus ensuring the data security of edge devices.

[0030] 2. This invention employs MPC or federated learning technology for cross-system data interaction. The original data remains locally, and a blockchain audit log is generated to record the calling behavior. This achieves data sharing while ensuring that the data does not leave the local machine, protecting data privacy and providing a basis for tracing data calls.

[0031] 3. This invention calculates permission levels based on a risk assessment matrix, combined with user roles, device health, and network parameters. When network latency is high or water levels exceed warning thresholds, remote control permissions are automatically reduced to read-only to prevent accidental operation under abnormal circumstances and ensure the safety of water conservancy facility control. Attached Figure Description

[0032] Figure 1 This is a flowchart of the method of the present invention;

[0033] Figure 2 This is a tree diagram of the system modules of the present invention;

[0034] Figure 3 This is the dynamic permission evaluation tree of the present invention;

[0035] Figure 4 This is a timing diagram of data fragmentation transmission according to the present invention;

[0036] Figure 5 This is the key management state diagram of the present invention. Detailed Implementation

[0037] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and should not be construed as limiting the scope of the invention.

[0038] A water conservancy information security protection method and system includes the following steps:

[0039] S1 edge device physical security hardening: Deploy physical security protection modules in hydraulic sensors and RTU remote terminal units to monitor the vibration, open cover status and environmental parameters of the equipment in real time. When abnormal physical attacks or environmental limits are detected, the key is automatically erased and the data protection mode is activated.

[0040] The physical security reinforcement step of the edge device achieves active physical layer protection of the water conservancy terminal equipment by integrating multimodal sensors and a security control unit. The physical security protection module consists of a high-sensitivity triaxial accelerometer, a digital temperature and humidity sensor, a dual-frequency GPS positioning module, and a triaxial tilt sensor. The accelerometer has a range of ±20g and a resolution of 0.1g. It collects the vibration spectrum characteristics of the equipment in real time and compares them with a preset security baseline. When the vibration energy is detected to exceed the threshold for 2 seconds in the 50-200Hz frequency band or when the Hall effect cover opening detection circuit is triggered for more than 1 second, it is determined to be a physical attack event. At the same time, an event fingerprint containing the attack time, vibration spectrum characteristics, and GPS coordinates is generated and written to the anti-tampering log area after SM3 hash operation.

[0041] Furthermore, for monitoring environmental anomalies, the safe temperature range is set to -20℃ to +60℃, the GPS position offset threshold is ≥5 meters, and the tilt angle allowable deviation is ±2°. When the temperature exceeds the limit for 5 minutes without recovery, the position offset exceeds the threshold, or the tilt angle is abnormal, the environmental over-limit response protocol is activated.

[0042] The security control unit has a built-in trusted execution environment based on the ARM Cortex-M4 architecture. After detecting a physical attack or environmental exceedance, it erases the encrypted master key and session key stored in the TEE within 0.3 seconds, and transfers the untransmitted encrypted data to the physical write-protected NOR Flash memory. At the same time, it sends encrypted alarm information to the monitoring center through the Beidou RDSS short message service. The alarm content includes the device ID, event type and WGS-84 geographic coordinates.

[0043] In addition, to improve reliability, an improved Kalman filter algorithm is used to process vibration signals to distinguish between mechanical shocks and malicious attacks. Polynomial fitting compensation is used to compensate for temperature drift to eliminate interference from slow environmental changes. A dual power supply redundancy mechanism and a three-mode redundancy architecture for key sensors are deployed. The final monitoring results are output through a majority voting mechanism to ensure the continuity of protection functions under extreme weather conditions.

[0044] S2 Dynamic Encrypted Data Acquisition: Based on the PUF Physical Unclonable Function, a unique fingerprint of the device is generated. Combined with environmental noise, an encryption key is dynamically generated. The collected water conservancy monitoring data is encrypted after adding a triplet identifier of device fingerprint, timestamp and geographic coordinates. The key validity period is set to 3-10 minutes.

[0045] S3 Adaptive Fragmented Redundant Transmission: The number of data fragments N≥3 is dynamically determined based on network latency and packet loss rate. The main transmission path is encrypted using the national cryptographic SM4 algorithm, while the backup path uses a lightweight encryption algorithm. The receiving end verifies data integrity through hash verification and blockchain evidence storage.

[0046] S4 Blockchain - Distributed Consortium Storage: Data hash values ​​are written to consortium blockchain nodes, the original data is encrypted and stored in fragments on the IPFS interplanetary file system, and overdue data is destroyed through smart contracts;

[0047] S5 privacy and security data sharing: When interacting with data across systems, it uses MPC secure multi-party computation or federated learning technology to ensure that the original data does not leave the local storage node and generates a blockchain audit log to record data call behavior;

[0048] The privacy-preserving data sharing steps achieve secure collaboration of cross-system data by integrating cryptographic protocols and distributed computing frameworks. Specifically, they include the following progressive technical solutions: In cross-basin or cross-departmental data interaction scenarios, to address the privacy protection needs of hydrological monitoring raw data, a privacy intersection protocol based on secure multi-party computation or a vertical federated learning architecture is adopted. The MPC protocol implements additive homomorphic encryption based on Shamir secret sharing and Beaver triple pre-computation. Participants exchange encrypted intermediate values ​​through a ring topology structure. Local data is encrypted by the Paillier algorithm and then divided into multiple secret fragments. The number of fragments and the number of participants satisfy N≥2t+1, where t is the fault tolerance threshold.

[0049] Each fragment is signed using the national cryptographic SM2 algorithm and then transmitted to the collaborating party via a quantum key distribution channel. In the federated learning framework, the global model parameters are updated through homomorphic encrypted gradients to ensure the aggregability of gradients and the irreversibility of plaintext. Data users must pass identity authentication based on zero-knowledge proofs to prove that they have access rights in accordance with the "Water Conservancy Data Sharing Management Measures" and that the purpose of data use is limited to flood control scheduling or engineering safety assessment.

[0050] All data interaction behaviors trigger the generation of blockchain audit logs in real time. The log content includes data hash values, caller digital certificate fingerprints, operation timestamps, and data usage declarations. The log structure is organized using MerklePatriciaTree and written to consortium blockchain nodes. The block generation interval is dynamically adjusted according to the urgency of water conservancy business, set to 5 minutes / block during the flood season and extended to 1 hour / block during the non-flood season. To cope with network partitions or node failures, an asynchronous Byzantine fault-tolerant consensus mechanism is deployed to ensure that the audit log can still achieve consistency when 1 / 3 of the nodes fail. For highly sensitive data, differential privacy (DP) protection is additionally enabled through Laplace noise injection, where the noise scale ε = 0.5. After Laplace noise injection, statistical features are obfuscated to prevent member inference attacks.

[0051] The compliance of the entire data sharing process is automatically verified by smart contracts. The contract code solidifies the terms of the "Hydrological Data Use License Agreement". When an unauthorized data export attempt is detected, the session is immediately terminated and a security alarm event is triggered. The alarm information is stored in IPFS shards and then linked to the blockchain traceability index.

[0052] S6 Multi-Dimensional Threat Real-Time Handling: Based on the user's historical operation sequence, a behavior baseline model is trained to detect abnormal access behavior in real time. When it is identified that an overseas IP is logging in outside of working hours or scanning high-risk ports, it will link with network security devices to generate an Access Control List (ACL) for millisecond-level blocking.

[0053] S7 Dynamic Permission Intelligent Control: Establishes a risk assessment matrix, calculates dynamic permission levels by comprehensively considering user role permissions, equipment health indicators, and network environment parameters. When the network latency of the reservoir monitoring point is greater than 200ms or the water level exceeds the warning line, the remote control command permission is automatically downgraded to read-only mode.

[0054] The dynamic permission intelligent control step achieves real-time adaptive adjustment of water conservancy system access permissions by constructing a multi-dimensional risk assessment matrix. The risk assessment matrix uses user role permissions as the basic weight factor, combined with equipment health quantification indicators and network environment dynamic parameters for comprehensive calculation. Specifically, user role permissions are divided into ordinary query personnel (weight coefficient 0.2), data administrators (0.4), and system controllers (0.6). Equipment health indicators are calculated based on sensor vibration velocity (threshold ≤ 4.5 mm / s), equipment temperature rise (gate motor ≤ 40K), and power supply voltage fluctuation range (±10% of rated value) to calculate a normalized health score Sdevice∈[0,1]. Network environment parameters are generated based on real-time network latency (baseline value ≤ 200ms), packet loss rate (≤ 5%), and Beidou signal strength (≥ -90dBm) to generate an environmental risk coefficient Nenu.

[0055] The risk assessment value, Risk, is dynamically calculated using the linear weighted formula Risk = 0.5Puser + 0.3Sdevice + 0.2Nenu. When Risk ≥ 7 (out of 10) or a preset water conservancy business threshold is triggered, the remote control command permission is automatically downgraded to read-only mode, prohibiting the issuance of commands for gate opening adjustment, pump station start / stop, and flood discharge. The permission control engine connects to the water conservancy business system in real time. When the dam displacement rate ≥ 2mm / h or the irrigation canal flow exceeds the design value by 15m³, the remote control command will be activated. 3 At / s, force an increase in the risk assessment frequency to once per second and enable multi-source data cross-validation;

[0056] After the permission status change information is signed by the SM2 algorithm, it is written into the blockchain audit chain and synchronously updated to the distributed access control list. After the network recovers and the water level falls below the security threshold, the control permissions are gradually restored according to the manual review process.

[0057] Meanwhile, to cope with extreme network outage scenarios, a local edge decision unit is deployed. Based on the LSTM model, the device health trend is predicted. When the vibration speed is predicted to exceed 4.0 mm / s in the next 10 minutes, the permission downgrade is pre-triggered and the control command is cached to the local trusted execution environment. After the network is restored, the delayed operation is executed after dual signature authentication to ensure a balance between business continuity and security.

[0058] Example 1:

[0059] Dam safety monitoring scenario;

[0060] The application scenario is the safety protection of sensors at surface displacement monitoring points of a concrete gravity dam;

[0061] The technical implementation involves adding a physical security hardening step for edge devices to the traditional data acquisition process;

[0062] Operation procedure: When the dam displacement sensor detects a displacement of ≥20mm in the X direction, the anti-tamper chip is triggered to erase the key;

[0063] Then, an alarm is sent to the monitoring center via BeiDou short message to block the upload of abnormal data. See Table 1 for specific parameters.

[0064] Table 1:

[0065]

[0066] Example 2:

[0067] Cross-regional reservoir scheduling scenario

[0068] Application scenario: Data sharing for joint flood control scheduling of reservoir groups across provinces and river basins;

[0069] Technical implementation: Adding privacy-preserving computation to the traditional data sharing process for data sharing;

[0070] Operating procedures:

[0071] When the water level of the reservoir in Province A exceeds 183m, the optimal flood discharge plan is calculated through MPC safety multi-party calculation and joint calculation with the two neighboring provinces.

[0072] Furthermore, the original water level data does not leave the local machine; only encrypted intermediate values ​​are exchanged. Finally, the calculation results are stored on the blockchain for evidence. Specific parameters are shown in Table 2.

[0073] Table 2:

[0074]

[0075] Example 3:

[0076] Remote control system scenario for gates;

[0077] Application scenario: Dynamic management of remote control permissions for irrigation canal gates;

[0078] Technical Implementation:

[0079] Add dynamic permission adaptive adjustment to traditional access control;

[0080] When the network latency reaches 250ms and the channel water level is close to 2.6m;

[0081] Furthermore, the system automatically downgrades remote control permissions to "read-only mode";

[0082] Then, the gate opening adjustment command is disabled;

[0083] Only water level data queries are allowed at this time;

[0084] See Table 3 for specific parameters;

[0085] Table 3:

[0086]

[0087] Example 4:

[0088] Integrated scenario of flood control command system;

[0089] It employs all the methods and steps described above in the technical solution;

[0090] Application scenario: Multi-source data fusion platform for provincial flood control and drought relief command centers;

[0091] Operating procedures:

[0092] First, the rain gauge data is encrypted using PUF. After the key expires in 5 minutes, it is transmitted to the command center in three separate streams. Then, the hydrological data is stored in the IPFS network.

[0093] Among them, the flood prediction model is trained through federated learning;

[0094] In addition, finally, when abnormal access to satellite cloud image data by overseas IP addresses is detected;

[0095] This fact shows that the DRL model blocks the connection and initiates permission backtracking review within 300ms. Moreover, with its real-time monitoring and intelligent analysis capabilities, the system can make a quick judgment after identifying potential security risks or violations. Specific parameter names and data ranges are shown in Table 4.

[0096] Table 4:

[0097]

[0098] In summary, the technical solution of this invention addresses the issue of water conservancy information security by innovating in multiple scenarios. In the dam safety monitoring scenario, physical security reinforcement is implemented for edge devices; in the cross-regional reservoir scheduling scenario, privacy-preserving data sharing technology is employed.

[0099] The remote gate control scenario enables dynamic and intelligent access control. By integrating various innovations into the flood control command system, a comprehensive security protection system is constructed for the entire process of data collection, transmission, storage, and decision-making. Through technological integration, a physical-logical-access protection architecture is formed, covering water conservancy safety standards, reducing system operation and maintenance costs, improving system security and business processing efficiency, meeting the network security requirements of smart water conservancy, and providing security guarantees for water conservancy project construction.

[0100] All embodiments of the present invention are within the scope of protection of this patent.

[0101] The embodiments of the present invention are given for the purposes of illustration and description. Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A water conservancy information security protection method, characterized in that, The method comprises the following steps: S1 Physical security reinforcement of edge devices: A physical security protection module is deployed in water conservancy sensors and RTU remote terminal units to monitor the vibration, cover opening state and environmental parameters of the devices in real time. When an abnormal physical attack or environmental overrun is detected, the key is automatically erased and the data protection mode is started; S2 Dynamic encryption data collection: A device unique fingerprint is generated based on a PUF physical unclonable function, and an encryption key is dynamically generated in combination with environmental noise. The collected water conservancy monitoring data is encrypted after being identified by a three-tuple of device fingerprint, timestamp and geographic coordinates. The key validity period is set to 3-10 minutes; S3 Self-adaptive sharding redundant transmission: The number of data shards N≥3 is dynamically determined according to network delay and packet loss rate. The main transmission path is encrypted using the SM4 algorithm, the backup path is encrypted using a lightweight encryption algorithm, and the receiving end verifies the data integrity through hash checking and blockchain storage; S4 Blockchain-distributed joint storage: The data hash value is written to the alliance chain node, and the original data is stored in the IPFS interstellar file system after being encrypted and sharded, and the intelligent contract is used to realize the targeted destruction of data beyond the period; S5 Privacy-safe data sharing: In cross-system data interaction, MPC secure multi-party computing or federated learning technology is used to ensure that the original data does not leave the local storage node, and a blockchain audit log is generated to record data calling behavior; S6 Real-time disposal of multi-dimensional threats: A behavior baseline model is trained based on user historical operation sequences to detect abnormal access behavior in real time. When an overseas IP non-working time login or high-risk port scanning is identified, the network security device generates an access control list ACL for millisecond-level blocking; S7 Dynamic permission intelligent regulation: A risk assessment matrix is established to calculate the dynamic permission level by combining user role permissions, device health index and network environment parameters. When the network delay of the reservoir monitoring point is >200ms or the water level exceeds the warning line, the remote control instruction permission is automatically downgraded to read-only mode; The generation rule of the dynamic encryption key is that the key validity period is dynamically adjusted according to the water conservancy monitoring parameter, the greater the difference between the water level and the warning water level, the shorter the key validity period, the encryption data packet header is attached with water conservancy engineering metadata, including the monitoring point elevation, data collection time and equipment installation coordinates, wherein the key validity period is dynamically associated with the water conservancy monitoring parameter, and satisfies the formula: = .

2. The water information security protection method of claim 1, wherein The determination conditions of the abnormal physical attack include vibration frequency ≥50Hz, corresponding acceleration ≥5g or device shell cover opening lasting more than 2 seconds, and the environmental parameter overrun includes temperature exceeding the range of-20℃~+60℃, humidity >90%RH or GPS position offset ≥5 meters.

3. The water information security protection method of claim 1, wherein When the dam displacement rate reaches 2mm / hour and above, the key validity period will be shortened to 50% of the normal state. The encrypted data packet header is attached with water conservancy engineering metadata, including monitoring point elevation, data collection time and device installation coordinates.

4. The water information security protection method of claim 1, wherein, The number of shards is dynamically determined according to network quality. The higher the network delay, the more shards. When the network delay exceeds 300ms, the Beidou short message is forced to be used as a backup transmission path. The main path uses the SM4-GCM encryption mode, the backup path uses the ChaCha20-Poly1305 algorithm, and the keys are independently generated and not reused.

5. The water information security protection method of claim 1, wherein, The user risk level of the risk assessment matrix is divided into 1-3 levels, 1 level is a general inquirer, and 3 level is a system administrator, the equipment health degree is calculated according to the vibration speed of the equipment, the greater the vibration speed, the lower the equipment health degree, when the vibration speed exceeds a certain threshold, the equipment health degree is zero, the network environment risk coefficient is 1 when the network delay exceeds 200 milliseconds, otherwise it is 0; when the risk assessment result reaches 6 and above, the flood discharge gate opening adjustment, pump station unit start-stop high-risk operation is prohibited.

6. The water information security protection method of claim 1, wherein, When the gate motor temperature rise relative to the ambient temperature reaches 40K, or the power supply voltage fluctuates more than ±20% based on the rated 24VDC, the equipment health degree is directly zero, and when the reservoir water level exceeds the design flood level, the network environment risk coefficient will be forced to be the maximum value 1.

7. The water conservancy information security protection system of claim 1, wherein, The following modules are included: Physical security awareness module: deployed in water conservancy terminal equipment, integrated with anti-disassembly machine self-destruction unit and environmental sensor, real-time monitoring of equipment physical state; Dynamic encryption acquisition module: generate device unique fingerprint through PUF physical unclonable function, dynamically generate encryption key combined with environmental noise, encrypt the collected water conservancy engineering data; Adaptive fragmentation transmission module: according to network delay, packet loss rate and water conservancy business priority, the encrypted data is fragmented and transmitted through multiple paths, supporting dynamic switching of national encryption SM4 and ChaCha20 algorithm; Distributed storage and evidence module: composed of blockchain nodes and IPFS storage cluster, the blockchain records data hash value and operation log, IPFS stores encrypted fragments and performs intelligent contract driven life cycle management; Privacy computing and sharing module: based on MPC secure multi-party computing and federated learning framework, realize cross-regional water conservancy data joint analysis, original data does not leave local storage node; Intelligent permission control module: built-in risk assessment matrix and user behavior baseline model, dynamically adjust user operation permission, and execute real-time blocking strategy in cooperation with firewall.

8. The water information security protection system of claim 7, wherein, The water conservancy terminal equipment includes water level gauge, displacement sensor, gate controller, and the flood control instruction in the water conservancy business is the highest level.

Images