An evaluation method and system for port ecological construction

Abstract

This invention relates to the technical field of port ecological construction, specifically disclosing a port ecological construction assessment method and system. The method comprises the following steps: collecting dynamic ecological and environmental data from port construction and operation using a multi-source sensor array; performing spatiotemporal coupled analysis of the data, using an adaptive sliding time window algorithm to calculate the dynamic attenuation rate of environmental indicators in various dimensions (#imgabs0#); adjusting the ecological response time constant θ based on the hydrodynamic conditions of the port's surrounding waters; synchronously integrating and calculating the historical cumulative effect (#imgabs1#) to generate an environmental health evolution surface; constructing a comprehensive assessment model; and predicting environmental risk values based on a hybrid prediction model. This invention provides an intelligent, automated, safe, reliable, green, and low-carbon assessment solution for port ecological construction, significantly improving port ecological protection and sustainable development.

Description

Technical Field

[0001] The present invention relates to the technical field of port ecological construction, and in particular to a port ecological construction evaluation method and system. Background Art

[0002] With the rapid development of the global port economy, port ecological construction has become a key link in achieving sustainable development. Traditional port construction and operation processes are faced with problems such as fragmented ecological environment monitoring data, insufficient adaptability of dynamic assessment models, weak multi-objective decision support capabilities, and the lack of data security and evidence storage mechanisms. These problems are specifically manifested as follows:

[0003] 1. Data collection and processing: Existing systems rely on single sensors or local monitoring networks, making it difficult to comprehensively acquire multi-dimensional dynamic ecological and environmental data such as water quality, air, soil, and biodiversity. Furthermore, they lack efficient multimodal data fusion and trusted evidence storage technologies, limiting data integrity and reliability.

[0004] 2. Assessment model level: Traditional assessment methods often use static weight allocation and fixed time windows. These methods are unable to respond in real time to the spatiotemporal heterogeneity of the port environment (such as tidal changes and differences in hydrodynamic conditions). They also lack the precision to quantify ecological cumulative effects and dynamic attenuation processes, making it difficult to accurately generate trends in environmental health evolution.

[0005] 3. Decision-making support: There is a lack of intelligent optimization algorithms that integrate multiple objectives, such as carbon footprint and ecological connectivity. Predictive models do not fully incorporate port operational characteristics (such as ship scheduling and resource utilization efficiency), resulting in poorly targeted and cost-effective assessment reports and construction plans. Furthermore, traditional systems lack the integration of technologies such as digital twins and augmented reality, making it difficult for decision makers to intuitively perceive ecological impacts and providing a limited interactive decision-making experience.

[0006] 4. Data security: Sensor data is vulnerable to tampering or attacks during transmission and storage. Traditional encryption technologies cannot meet the collaborative verification needs of multiple port stakeholders (management departments, ecological and environmental protection agencies, and enterprises), and there is a lack of a balanced mechanism between data sharing and privacy protection.

[0007] In summary, the existing port ecological construction assessment technology has significant shortcomings in data reliability collection, dynamic modeling, multi-objective optimization and intelligent decision-making. A systematic solution that integrates advanced information technology and ecological assessment theory is needed. Summary of the Invention

[0008] In view of the deficiencies in the prior art, the present invention provides a method and system for evaluating port ecological construction, which solves the problems in the above-mentioned background technology.

[0009] To achieve the above objectives, the present invention is implemented through the following technical solutions: A port ecological construction evaluation method includes the following steps:

[0010] S1. Multi-source data collection and trusted evidence storage:

[0011] A multi-source sensor array is used to collect dynamic ecological and environmental data during port construction and operation, including water quality parameter groups, air pollutant concentrations, soil heavy metal content, noise spectrum, biodiversity index, and resource utilization efficiency indicators. A multimodal feature fusion matrix is constructed, and the original data is credibly stored.

[0012] S2. Spatiotemporal dynamic analysis and environmental health assessment:

[0013] The data is analyzed in time and space, and the dynamic attenuation rate of environmental indicators in each dimension is calculated using an adaptive sliding time window algorithm. , adjust the ecological response time constant θ based on the hydrodynamic conditions of the sea area around the port, and synchronously calculate the historical cumulative effect by integration , generate the environmental health evolution surface;

[0014] S3. Construction of comprehensive evaluation model:

[0015] A comprehensive assessment model was constructed to quantify the habitat fragmentation index and the ecological connectivity index (ECON). A dynamic weight allocation algorithm was used to integrate subjective and objective evaluation factors. An adaptive learning mechanism was introduced to optimize the weight convergence process, and the output was the ecological comprehensive index (ECI).

[0016] S4. Environmental risk prediction and decision support:

[0017] The environmental risk value is predicted based on the hybrid prediction model, and the Pareto frontier solution set is generated through the multi-objective optimization algorithm. Combined with the port ecological construction goals, a graded assessment report and construction plan set are generated, and the prediction results are fed back to the digital twin engine for dynamic calibration.

[0018] Preferably, the credible evidence of original data in S1 includes:

[0019] Blockchain technology is used to hash and timestamp sensor data on the chain to build an unalterable data source proof chain;

[0020] The blockchain evidence storage system adopts a consortium chain architecture, with nodes including port management departments, ecological and environmental protection agencies and monitoring equipment manufacturers, and verifies data integrity through a consensus mechanism;

[0021] A quantum random number generator is introduced to generate data encryption keys, and quantum key distribution technology is used to ensure the security of key transmission and prevent data from being illegally cracked.

[0022] Preferably, the dynamic decay rate in S2 Calculated by the following formula:

[0023] ;

[0024] in, is the dynamic time window period (4h≤τ(t)≤24h), is the adaptive attenuation coefficient, For the Always monitor the value, is the lower safety threshold of the environmental indicator after dynamic correction, It is the upper limit threshold of the environmental indicator after dynamic correction. The threshold is automatically calibrated by Monte Carlo simulation and real-time data deviation. is the exponential decay factor, Indicates the current time and historical moments the time interval between

[0025] Combined with the environmental variables of sea temperature and salinity obtained from satellite remote sensing data, a multivariate regression model is constructed to dynamically adjust the adaptive attenuation coefficient. , improve the accuracy of calculation of environmental indicator attenuation rate;

[0026] Introducing reinforcement learning algorithms to optimize dynamic time window cycles based on historical evolution trends of environmental indicators and real-time monitoring data adjustment strategy.

[0027] Preferably, the dynamic weight allocation algorithm in S3 is embedded in the attention mechanism, and the weights of key indicators are calculated as follows:

[0028] ;

[0029] in, is the attention weight of the key indicator, 、 For the feature extraction function, the weight matrix is dynamically optimized by minimizing the evaluation error;

[0030] At the same time, knowledge graph technology is introduced to build a knowledge graph in the port ecological field to assist in adjusting weight distribution; and the federated learning framework is combined to optimize the feature extraction function 、 , improving the generalization ability of the model.

[0031] Preferably, the hybrid prediction model in S4 is an improved STL-LSTM model, including the following features:

[0032] Hampel filter is used to eliminate outliers in time series data, and seasonal decomposition is used to extract the periodic characteristics of environmental data;

[0033] Construct a spatiotemporal attention LSTM network and weight the port operation intensity index and tidal cycle parameters when updating the hidden layer state;

[0034] Introducing the carbon footprint correction term, the formula is:

[0035] ,in is the carbon balance index;

[0036] Integrating a generative adversarial network, the generator generates potential change scenarios of environmental risks, and the discriminator compares the predicted results with the generated scenarios to optimize the prediction accuracy of the STL-LSTM model;

[0037] Edge computing nodes are used to pre-process and preliminarily predict real-time data at the port site, and key feature data is transmitted to the cloud for in-depth analysis, reducing data transmission delays and cloud computing pressure.

[0038] A port ecological construction assessment system includes: an intelligent perception layer, a digital twin engine, a decision optimization layer and an augmented reality terminal; the intelligent perception layer deploys a multispectral water quality sensor array, a MEMS gas sensor network, and an underwater acoustic Doppler profiler to form a heterogeneous data acquisition matrix, and integrates a blockchain evidence storage module to realize trusted data on-chain; the digital twin engine integrates a spatiotemporal coupling analytical model and a hybrid prediction model, supports port environmental degree of freedom simulation, and includes a dynamic grid optimization unit and an ecological stress analysis unit; the decision optimization layer runs a dynamic weight allocation algorithm and a multi-objective optimization algorithm to generate a Pareto solution set three-dimensional decision cloud map, and embeds a carbon footprint tracking module to quantify the port's full-cycle carbon emissions and carbon sinks; the augmented reality terminal uses light field display technology to present the pollutant migration path, and combines tactile feedback and neural feedback devices to realize interactive adjustment of the plan.

[0039] Preferably, the blockchain evidence storage module includes: a data preprocessing unit and a smart contract unit; the data preprocessing unit denoises, normalizes and quality checks the sensor data; the smart contract unit automatically verifies the integrity of the data, and triggers an on-chain warning when the monitoring value deviates from the historical mean by more than 20%, and marks the data abnormal status bit; at the same time, zero-knowledge proof technology is introduced to verify the authenticity and integrity of the data to a third party without leaking the original data content, and shard storage technology is used to divide the blockchain data into multiple shards, which are stored on different nodes to improve data storage and retrieval efficiency.

[0040] Preferably, the dynamic grid optimization unit of the digital twin engine uses non-uniform rational B-splines and NURBS surface models to reconstruct the port terrain. The grid resolution is dynamically adjusted according to the ship's draft depth and tidal height, with a minimum resolution of 0.5 m × 0.5 m.

[0041] Incorporating point cloud processing technology, using port point cloud data acquired by drones and underwater robots, the NURBS surface model is updated in real time to improve the accuracy of the digital twin model.

[0042] The ecological stress analysis unit introduces a physical engine to simulate the impact of the physical environment of port water flow and wind on ships and facilities, achieving more realistic port operation scenario simulation.

[0043] Preferably, the neurofeedback evaluation module of the augmented reality terminal includes: an EEG signal acquisition device and a cortical activation map analysis unit; the EEG signal acquisition device obtains the decision maker's alpha and beta wave characteristics in real time; the cortical activation map analysis unit identifies the cognitive matching degree of the scheme through a convolutional neural network, and outputs optimization suggestions to the decision optimization layer; at the same time, emotional computing technology is introduced to analyze the emotional characteristics of the decision maker's facial expressions and voice intonation, and the EEG signals are combined to evaluate the decision maker's emotional preference for different construction schemes. The virtual scene generated by the digital twin engine is combined with virtual reality technology to allow the decision maker to immersively experience the changes in the port ecological environment under different construction schemes, thereby enhancing the intuitiveness and accuracy of the decision.

[0044] Preferably, the carbon footprint tracking module is used to calculate the carbon balance index of port operations. The specific calculation formula is as follows:

[0045] ;

[0046] in:

[0047] Represents ecological carbon sink, which is calculated based on wetland area and vegetation carbon sequestration efficiency;

[0048] Represents carbon emissions during the operation phase, which is estimated based on energy consumption and carbon emission factors;

[0049] represents the embodied carbon emissions during the construction phase. All three are included in the constraints of the multi-objective optimization model;

[0050] At the same time, the life cycle assessment method is introduced to conduct carbon footprint analysis on the entire life cycle of port construction, operation, maintenance to abandonment, so as to more comprehensively evaluate the carbon emissions of the port; combined with carbon trading market data, a carbon cost assessment model is established to link the carbon balance index with economic costs, providing economic feasibility analysis for the port's ecological construction plan.

[0051] The present invention provides a method and system for evaluating port ecological construction, which has the following beneficial effects:

[0052] 1. Improvement of data collection and trusted evidence storage capabilities:

[0053] 1. By deploying arrays of multispectral water quality sensors and MEMS gas sensors, a dynamic data collection matrix covering environmental, resource, and biological dimensions is constructed. Combined with multimodal feature fusion technology, comprehensive perception and real-time monitoring of port ecological elements is achieved.

[0054] 2. Adopting alliance chain architecture and quantum key distribution technology, sensor data is hashed, timestamped and dynamically key managed to ensure that data cannot be tampered with and transmission is secure, providing a trusted data source for collaborative verification among multiple parties and solving the reliability issues of data storage in traditional systems.

[0055] 2. Breakthrough in dynamic assessment and prediction accuracy:

[0056] 1. Based on an adaptive sliding time window and reinforcement learning algorithm, the time window period and attenuation coefficient are dynamically adjusted. The ecological response process is quantified based on the hydrodynamic conditions of the sea area surrounding the port. The historical cumulative effects of environmental indicators are accurately calculated, and a high-resolution environmental health evolution surface is generated to enhance the spatiotemporal dynamic assessment capability.

[0057] 2. By embedding attention mechanisms and knowledge graph technology, the system dynamically integrates subjective and objective evaluation factors to achieve adaptive optimization of the weights of key indicators such as habitat fragmentation and ecological connectivity. By introducing carbon footprint correction terms and generative adversarial networks through an improved STL-LSTM hybrid prediction model, the accuracy of environmental risk prediction is significantly improved, providing a scientific and quantitative basis for the ecological construction of ports.

[0058] 3. Decision support and interactive experience innovation:

[0059] 1. Utilizing NURBS surface models and point cloud processing technology, a high-precision digital twin engine is constructed to dynamically simulate physical environments such as port terrain, water flow, and wind. Combined with the Pareto frontier solution set, a three-dimensional decision cloud map is generated to enable decision makers to intuitively assess the ecological impact of different options.

[0060] 2. Through EEG signal acquisition, affective computing technology, and virtual reality scenarios, immersive interaction between decision makers and the evaluation system can be achieved, cognitive preferences and emotional feedback can be captured in real time, and personalized construction plans can be generated to improve decision-making efficiency and scientificity.

[0061] 4. Full-cycle low-carbon assessment and economic analysis:

[0062] 1. Construct a carbon balance index model that includes carbon emissions and ecological carbon sinks during the construction and operation phases. Combining life cycle assessment methods with carbon trading market data, the model links ecological assessments with economic costs, providing quantitative support for the port's low-carbon transformation and helping achieve dual carbon goals.

[0063] 2. Preprocess real-time data through edge nodes and transmit key features to the cloud, reducing data transmission latency and computing pressure, and improving system response speed and resource utilization efficiency.

[0064] 5. Technology Integration and System Scalability:

[0065] Constructing a dynamically adjustable intelligent evaluation framework supports the flexible expansion of the port ecological construction indicator system and model generalization, adapting to the geographical environment and development needs of different ports, and has broad application prospects and technological foresight.

[0066] In summary, the present invention provides an intelligent, automated, safe, reliable, green and low-carbon assessment solution for port ecological construction through full-chain technological innovation in data collection, analysis and modeling, intelligent decision-making, and interactive optimization, significantly improving the level of port ecological protection and sustainable development. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 This is a flow chart of a method for evaluating port ecological construction according to the present invention;

[0068] Figure 2 This is a principle block diagram of a port ecological construction assessment system of the present invention;

[0069] Figure 3 This is the principle block diagram of the digital twin engine;

[0070] Figure 4 This is the principle block diagram of the blockchain evidence storage module;

[0071] Figure 5 This is the principle block diagram of the augmented reality terminal. DETAILED DESCRIPTION

[0072] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0073] like Figure 1 As shown, the present invention provides a technical solution: a port ecological construction evaluation method, comprising the following steps:

[0074] S1. Multi-source data collection and reliable evidence storage: Dynamic ecological and environmental data from port construction and operation are collected through a multi-source sensor array. The data includes water quality parameter groups, air pollutant concentrations, soil heavy metal content, noise spectrum, biodiversity index, and resource utilization efficiency indicators. A multimodal feature fusion matrix is constructed, and the original data is reliably stored.

[0075] S2. Spatiotemporal dynamic analysis and environmental health assessment: Perform spatiotemporal coupled analysis on the data and use an adaptive sliding time window algorithm to calculate the dynamic attenuation rate of environmental indicators in each dimension. , adjust the ecological response time constant θ based on the hydrodynamic conditions of the sea area around the port, and synchronously calculate the historical cumulative effect by integration , generate the environmental health evolution surface;

[0076] S3. Construction of a comprehensive assessment model: Build a comprehensive assessment model to quantify the habitat fragmentation index and the ecological connectivity index (ECON). A dynamic weight allocation algorithm is used to integrate subjective and objective evaluation factors. An adaptive learning mechanism is introduced to optimize the weight convergence process, and the output is the ecological comprehensive index (ECI).

[0077] S4. Environmental risk prediction and decision support: Predict environmental risk values based on a hybrid prediction model, generate a Pareto frontier solution set through a multi-objective optimization algorithm, generate a graded assessment report and a construction plan set based on the port's ecological construction goals, and feed the prediction results back to the digital twin engine for dynamic calibration.

[0078] More specifically, the credible evidence of original data in S1 includes:

[0079] Blockchain technology is used to hash and timestamp sensor data on the chain to build an unalterable data source proof chain; the blockchain evidence storage system adopts a consortium chain architecture, with nodes including port management departments, ecological and environmental protection agencies and monitoring equipment manufacturers, and verifies data integrity through a consensus mechanism; a quantum random number generator is introduced to generate data encryption keys, and quantum key distribution technology is used to ensure the security of key transmission and prevent data from being illegally cracked.

[0080] The sensor uses a QRNG to generate quantum keys and encrypt the original data, ensuring that the data has not been tampered with (hash value matching) and the key cannot be leaked (QKD transmission). The encrypted data is transmitted through a hybrid quantum and classical channel, with QKD technology ensuring key security. Blockchain nodes verify data integrity through a consensus mechanism (hash value and timestamp matching). The data is stored in the consortium blockchain as the encrypted original text, hash chain, and timestamp. Any modification requires consensus from a majority of nodes and leaves an indelible record of the modification (such as adding a block to mark an anomaly).

[0081] More specifically, the dynamic decay rate in S2 Calculated by the following formula:

[0082] ;

[0083] in, is the dynamic time window period (4h≤τ(t)≤24h), is the adaptive attenuation coefficient, For the Always monitor the value, is the lower safety threshold of the environmental indicator after dynamic correction, The threshold is the upper limit of the safety of the environmental indicator after dynamic correction. The threshold is automatically calibrated with the deviation of real-time data through Monte Carlo simulation. The environmental variables of sea temperature and salinity obtained by satellite remote sensing data are combined to build a multivariate regression model and dynamically adjust the adaptive attenuation coefficient. , improve the accuracy of environmental indicator attenuation rate calculation; introduce reinforcement learning algorithm to optimize the dynamic time window period according to the historical evolution trend of environmental indicators and real-time monitoring data adjustment strategies; is the exponential decay factor, Indicates the current time and historical moments the time interval between Determines the size of the number attenuation factor, The larger the value, the more historical moments Distance from current time The farther, the The smaller the value of Monitoring value For the current moment The impact of environmental indicators on the decay rate is also weaker, reflecting the time lag effect of environmental impacts. For example, a port may discharge a certain amount of pollutants at a certain moment. Over time, these pollutants will diffuse and degrade in the environment. The longer the time since the discharge, the smaller the impact on the current environmental health.

[0084] in, is the normalized difference term, Monitoring value at all times Mapped to the interval [0, 1], it reflects the degree to which the indicator deviates from the safety lower limit at that moment (the larger the value, the farther it deviates from the safety threshold, and the greater the environmental pressure). is the lower safety threshold of the environmental indicator after dynamic correction, The upper safety threshold of the dynamically corrected environmental indicator is automatically calibrated with the deviation of real-time data through Monte Carlo simulation. is an exponential decay factor, reflecting the "time lag effect" of environmental impact. The longer the monitoring value is, the smaller its impact on the current moment is, and it decays exponentially over time ( The larger the value, the faster the decay rate). For example, if the noise level exceeds the standard at a certain moment, its impact on the current environmental health will gradually weaken over time (such as diffusion by seawater and dilution by air). The exponential function can better reflect the actual physical diffusion process. Dynamic time window averaging, in the length of The system averages the attenuated, normalized values within a dynamically adjustable time window (4-24 hours) to reflect the combined impact of recent environmental indicators. This avoids interference from outliers at a single moment, while adapting to cyclical changes in the port environment (such as tidal cycles and peak ship operations) through a dynamic window. A weighted average (heavier weights for recent data and less weights for more distant data) captures the dynamic attenuation of environmental indicators over time, generating the underlying data for the environmental health evolution surface. Dynamic adjustment of attenuation parameters based on port hydrodynamic conditions (such as tides and current velocity) addresses the inability of traditional fixed-window models to cope with rapidly changing environments.

[0085] More specifically, the dynamic weight allocation algorithm in S3 is embedded in the attention mechanism, and the weights of key indicators are calculated as follows:

[0086] ;

[0087] in, is the attention weight of the key indicator, 、 For the feature extraction function, the weight matrix is dynamically optimized by minimizing the evaluation error; at the same time, the knowledge graph technology is introduced to build a knowledge graph in the port ecological field to assist in adjusting the weight distribution; and the feature extraction function is optimized in combination with the federated learning framework. 、 , improving the generalization ability of the model.

[0088] Feature extraction function: and Feature extraction is performed on the Ecological Comprehensive Index (ECI) and the Biodiversity Comprehensive Index (BCI). These two functions map the original index data into a new feature space and extract feature information that has an important impact on weight distribution. For example May extract The characteristics related to energy consumption, pollutant emissions, etc. It may focus on characteristics such as the number of biological species and species distribution.

[0089] Attention function: Measured the The feature importance of the key indicators. and Perform some calculation (such as dot product, concatenation, etc.) to obtain a score that represents the importance of the indicator in the current evaluation.

[0090] Weight normalization: using the softmax function Normalize the attention score so that the sum of the weights of all key indicators is 1, so each It represents the The relative importance of each key indicator in the comprehensive evaluation.

[0091] A knowledge graph represents various entities in the port ecological domain (such as pollutants, biological species, and port facilities) and their relationships (such as pollution relationships and dependencies) in a graphical form. For example, a knowledge graph could include the causal relationship between oil spills and marine life mortality, or the correlation between sewage treatment facilities and water quality improvement. The process of constructing a knowledge graph includes steps such as data collection, entity identification, relationship extraction, and knowledge fusion. Knowledge can be obtained from a variety of data sources such as port monitoring data, scientific research literature, and industry standards. Natural language processing and machine learning techniques are then used to extract entities and relationships, and finally, this knowledge is integrated into a unified graph. The knowledge graph can provide prior knowledge and domain constraints for weight assignment. For example, when assessing the port ecological environment, if the knowledge graph shows that a certain pollutant has a serious impact on biodiversity, the weight of indicators related to that pollutant can be appropriately increased when calculating weights. Furthermore, the knowledge graph can help discover potential relationships between indicators, avoiding irrational weight assignment.

[0092] The federated learning framework optimizes feature extraction functions. Federated learning is a distributed machine learning technology that allows model training to be conducted across multiple participants (such as different port management departments and research institutions) without sharing raw data. In port ecological assessments, different ports may have different monitoring data and assessment experience. Federated learning can fully utilize these data resources and improve the model's generalization capabilities.

[0093] More specifically, the hybrid prediction model in S4 is an improved STL-LSTM model with the following features:

[0094] A Hampel filter is used to eliminate outliers in time series data, and seasonal decomposition is used to extract the periodic characteristics of environmental data. A spatiotemporal attention LSTM network is constructed, and the port operation intensity index and tidal cycle parameters are weighted when updating the hidden layer state.

[0095] Introducing the carbon footprint correction term, the formula is:

[0096] ,in The system uses a generative adversarial network (GAN) to generate scenarios for potential changes in environmental risks. The discriminator then compares the predicted results with the generated scenarios, optimizing the prediction accuracy of the STL-LSTM model. Edge computing nodes are used to preprocess and perform preliminary predictions on real-time port data, transmitting key feature data to the cloud for in-depth analysis, reducing data transmission latency and cloud computing pressure. During the dynamic monitoring and assessment of the port's ecological environment, data is collected chronologically, forming a time series. As a time marker, it can specifically represent a specific moment. For example, if data is collected in hours, =1 means the first hour, =2 means the second hour. This mark can clearly distinguish and correspond the environmental data and related calculation results at different times, making it easier to track and analyze the changing trends of the port's ecological environment over time.

[0097] in, It is the carbon balance index, which reflects the ratio of the port's ecological carbon sink to its full-cycle carbon emissions. is the correction coefficient (determined through historical data training, with the initial value set to 0.1, dynamically adjusting the weight of the impact of carbon balance on the prediction results). When the carbon sink is greater than the emission, the correction term is positive, improving the eco-friendliness of the prediction; otherwise, it depresses the prediction and quantifies the negative impact of carbon emissions on environmental risks. The carbon balance index directly links port carbon emissions to ecological protection goals, and the correction term makes the prediction results explicitly reflect the carbon neutrality orientation, providing a quantitative basis for low-carbon construction plans. Combining data preprocessing, spatiotemporal feature modeling, generative simulation, and distributed computing, a complete technical chain of data cleaning, feature enhancement, risk correction, and efficient deployment is constructed, significantly improving the robustness and practicality of the prediction model. The corrected environmental risk prediction value is more consistent with the actual port ecology.

[0098] like Figure 2-Figure 5As shown, a port ecological construction evaluation system includes: an intelligent perception layer, a digital twin engine, a decision optimization layer and an augmented reality terminal; the intelligent perception layer deploys a multispectral water quality sensor array, a MEMS gas sensor network, and an underwater acoustic Doppler profiler to form a heterogeneous data acquisition matrix, and integrates a blockchain evidence storage module to realize the trusted data on the chain; the digital twin engine integrates a spatiotemporal coupling analytical model and a hybrid prediction model, supports the simulation of the degree of freedom of the port environment, and includes a dynamic grid optimization unit and an ecological stress analysis unit; the decision optimization layer runs a dynamic weight allocation algorithm and a multi-objective optimization algorithm to generate a three-dimensional decision cloud map of the Pareto solution set, and embeds a carbon footprint tracking module to quantify the carbon emissions and carbon sinks of the port throughout the entire cycle; the augmented reality terminal uses light field display technology to present the migration path of pollutants, and combines tactile feedback and neural feedback devices to realize interactive adjustment of the plan.

[0099] Intelligent perception layer: Deployment of multispectral water quality sensor arrays (monitoring pH value, dissolved oxygen, and heavy metal ions), MEMS gas sensor networks (monitoring SO2, NOx, and VOCs), and underwater acoustic Doppler profilers (monitoring water flow velocity and direction), covering ecological data in six dimensions, including water quality, air, and hydrodynamics, to build a multimodal data acquisition network; the blockchain evidence storage module adopts a consortium chain architecture (nodes include port management, environmental protection agencies, and equipment manufacturers), and makes data tamper-proof through hash encryption and timestamps on the chain; integrated quantum key distribution technology ensures data transmission security, and combined with smart contracts to automatically verify data integrity, solving the problems of traditional data fragmentation and unreliable evidence.

[0100] More specifically, Figure 4 As shown in the figure, the blockchain evidence storage module includes: a data preprocessing unit and a smart contract unit; the data preprocessing unit denoises, normalizes and verifies the quality of sensor data; the smart contract unit automatically verifies the integrity of the data, triggers an on-chain warning when the deviation between the monitoring value and the historical mean exceeds 20%, and marks the data abnormal status bit; at the same time, the zero-knowledge proof technology is introduced to verify the authenticity and integrity of the data to a third party without leaking the original data content, and the sharding storage technology is used to divide the blockchain data into multiple shards and store them on different nodes to improve the efficiency of data storage and retrieval.

[0101] Digital Twin Engine: The spatiotemporal coupling analytical model calculates the dynamic attenuation rate of environmental indicators based on an adaptive sliding time window algorithm, quantifies the ecological cumulative effects in combination with hydrodynamic conditions, and generates an environmental health evolution surface; the hybrid prediction model embeds carbon footprint correction terms and generative adversarial networks to improve the accuracy of environmental risk prediction.

[0102] More specifically, Figure 3As shown, the dynamic grid optimization unit of the digital twin engine uses non-uniform rational B-splines and NURBS surface models to reconstruct the port terrain. The grid resolution is dynamically adjusted according to the ship's draft depth and tidal height, with a minimum resolution of 0.5 meters × 0.5 meters. Combined with point cloud processing technology, the port point cloud data obtained by drones and underwater robots is used to update the NURBS surface model in real time to improve the accuracy of the digital twin model. The ecological stress analysis unit introduces a physical engine to simulate the impact of the physical environment of port water flow and wind on ships and facilities, realizing a more realistic port operation scene simulation.

[0103] Decision optimization layer: The core is the dynamic weight allocation algorithm, which embeds the attention mechanism and knowledge graph, dynamically integrates subjective and objective factors such as habitat fragmentation and ecological connectivity, and outputs the ecological comprehensive index (ECI); generates the Pareto frontier solution set through the multi-objective optimization algorithm, balances ecological protection, economic costs and low-carbon goals, and outputs a three-dimensional decision cloud map (such as the three-dimensional relationship between ecological health, carbon emissions, and construction costs).

[0104] More specifically, the carbon footprint tracking module is used to calculate the carbon balance index of port operations. The specific calculation formula is as follows:

[0105] ;

[0106] in:

[0107] Represents ecological carbon sink, which is calculated based on wetland area and vegetation carbon sequestration efficiency;

[0108] Represents carbon emissions during the operation phase, which is estimated based on energy consumption and carbon emission factors;

[0109] represents the embodied carbon emissions during the construction phase. All three are included in the constraints of the multi-objective optimization model;

[0110] At the same time, the life cycle assessment method is introduced to conduct carbon footprint analysis on the entire life cycle of port construction, operation, maintenance to abandonment, so as to more comprehensively evaluate the carbon emissions of the port; combined with carbon trading market data, a carbon cost assessment model is established to link the carbon balance index with economic costs, providing economic feasibility analysis for the port's ecological construction plan.

[0111] More specifically, Figure 5As shown in the figure, the neural feedback evaluation module of the augmented reality terminal includes: an EEG signal acquisition device and a cortical activation map analysis unit; the EEG signal acquisition device obtains the decision maker's alpha and beta wave characteristics in real time; the cortical activation map analysis unit identifies the cognitive matching degree of the scheme through a convolutional neural network, and outputs optimization suggestions to the decision optimization layer; at the same time, the emotional computing technology is introduced to analyze the emotional characteristics of the decision maker's facial expressions and voice intonation, and the EEG signals are combined to evaluate the decision maker's emotional preference for different construction schemes. The virtual scene generated by the digital twin engine is combined with virtual reality technology to allow the decision maker to immersively experience the changes in the port ecological environment under different construction schemes, thereby enhancing the intuitiveness and accuracy of the decision.

[0112] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A method for evaluating port ecological construction, characterized in that: The following steps are involved: S1. Multi-source data collection and trusted evidence storage: A multi-source sensor array is used to collect dynamic ecological and environmental data during port construction and operation, including water quality parameter groups, air pollutant concentrations, soil heavy metal content, noise spectrum, biodiversity index, and resource utilization efficiency indicators. A multimodal feature fusion matrix is constructed, and the original data is credibly stored. S2. Spatiotemporal dynamic analysis and environmental health assessment: The data is analyzed in time and space, and the dynamic attenuation rate of environmental indicators in each dimension is calculated using an adaptive sliding time window algorithm. , adjust the ecological response time constant θ based on the hydrodynamic conditions of the sea area around the port, and synchronously calculate the historical cumulative effect by integration , generate the environmental health evolution surface; S3. Construction of comprehensive evaluation model: A comprehensive assessment model was constructed to quantify the habitat fragmentation index and the ecological connectivity index (ECON). A dynamic weight allocation algorithm was used to integrate subjective and objective evaluation factors. An adaptive learning mechanism was introduced to optimize the weight convergence process, and the output was the ecological comprehensive index (ECI). S4. Environmental risk prediction and decision support: The environmental risk value is predicted based on the hybrid prediction model, and the Pareto frontier solution set is generated through the multi-objective optimization algorithm. Combined with the port ecological construction goals, a graded assessment report and construction plan set are generated, and the prediction results are fed back to the digital twin engine for dynamic calibration.

2. A port ecological construction evaluation method according to claim 1, characterized in that: The credible evidence of original data mentioned in S1 includes: Blockchain technology is used to hash and timestamp sensor data on the chain to build an unalterable data source proof chain; The blockchain evidence storage system adopts a consortium chain architecture, with nodes including port management departments, ecological and environmental protection agencies and monitoring equipment manufacturers, and verifies data integrity through a consensus mechanism; A quantum random number generator is introduced to generate data encryption keys, and quantum key distribution technology is used to ensure the security of key transmission and prevent data from being illegally cracked.

3. A method for evaluating port ecological construction according to claim 2, characterized in that: Dynamic decay rate as described in S2 Calculated by the following formula: ； in, is the dynamic time window period, 4h≤τ(t)≤24h, is the adaptive attenuation coefficient, For the Always monitor the value, is the lower safety threshold of the environmental indicator after dynamic correction, It is the upper limit threshold of the environmental indicator after dynamic correction. The threshold is automatically calibrated by Monte Carlo simulation and real-time data deviation. is the exponential decay factor, Indicates the current time and historical moments the time interval between Combined with the environmental variables of sea temperature and salinity obtained from satellite remote sensing data, a multivariate regression model is constructed to dynamically adjust the adaptive attenuation coefficient. , improve the accuracy of calculation of environmental indicator attenuation rate; Introducing reinforcement learning algorithms to optimize dynamic time window cycles based on historical evolution trends of environmental indicators and real-time monitoring data adjustment strategy.

4. A method for evaluating port ecological construction according to claim 3, characterized in that: The dynamic weight allocation algorithm described in S3 is embedded in the attention mechanism, and the weights of key indicators are calculated as follows: ； in, is the attention weight of the key indicator, 、 For the feature extraction function, the weight matrix is dynamically optimized by minimizing the evaluation error. For the The ecological comprehensive index of each assessment object is used to quantify the comprehensive performance of the object in ecological construction. For the A comprehensive biodiversity index for each assessment object is used to measure the biodiversity status of the area covered by the object; Is to measure the Quantitative indicators of the comprehensive performance of each assessment object in port ecological construction; To quantify the The biodiversity status of the area covered by the assessment object; At the same time, knowledge graph technology is introduced to build a knowledge graph in the port ecological field to assist in adjusting weight distribution; and the federated learning framework is combined to optimize the feature extraction function 、 , improving the generalization ability of the model.

5. A method for evaluating port ecological construction according to claim 4, characterized in that: The hybrid prediction model described in S4 is an improved STL-LSTM model, which includes the following features: Hampel filter is used to eliminate outliers in time series data, and seasonal decomposition is used to extract the periodic characteristics of environmental data; Construct a spatiotemporal attention LSTM network and weight the port operation intensity index and tidal cycle parameters when updating the hidden layer state; Introducing the carbon footprint correction term, the formula is: ,in, is the carbon balance index, As a sign of time, is the correction factor, Indicates the The monitoring values of environmental indicators or related quantitative indicators, Based on the improved STL-LSTM model Environmental indicators in The original prediction results at the moment reflect the model's basic prediction of future changes in environmental indicators; is the carbon balance deviation term; Integrating a generative adversarial network, the generator generates potential change scenarios of environmental risks, and the discriminator compares the predicted results with the generated scenarios to optimize the prediction accuracy of the STL-LSTM model; Edge computing nodes are used to pre-process and preliminarily predict real-time data at the port site, and key feature data is transmitted to the cloud for in-depth analysis, reducing data transmission delays and cloud computing pressure.

6. A port ecological construction evaluation system, which is applied to a port ecological construction evaluation method according to claim 5, characterized in that: include: The intelligent perception layer deploys a multispectral water quality sensor array, a MEMS gas sensor network, and an underwater acoustic Doppler profiler to form a heterogeneous data acquisition matrix. It also integrates a blockchain evidence storage module to ensure reliable data storage on the chain. The digital twin engine integrates a spatiotemporal coupling analytical model and a hybrid prediction model, supports port environment degree of freedom simulation, and includes a dynamic grid optimization unit and an ecological stress analysis unit; The decision optimization layer runs a dynamic weight allocation algorithm and a multi-objective optimization algorithm to generate a three-dimensional decision cloud diagram of the Pareto solution set, and embeds a carbon footprint tracking module to quantify the port's full-cycle carbon emissions and carbon sinks; The augmented reality terminal uses light field display technology to present the migration path of pollutants, and combines tactile feedback and neural feedback devices to achieve interactive adjustment of the plan.

7. A port ecological construction evaluation system according to claim 6, characterized in that: The blockchain evidence storage module includes: Data preprocessing unit, which performs denoising, normalization and quality verification on sensor data; Smart contract units automatically verify data integrity. When the monitored value deviates from the historical average by more than 20%, an on-chain warning is triggered and the data abnormal status is marked. At the same time, zero-knowledge proof technology is introduced to verify the authenticity and integrity of data to a third party without leaking the original data content. Sharded storage technology is used to divide blockchain data into multiple shards and store them on different nodes to improve data storage and retrieval efficiency.

8. A port ecological construction evaluation system according to claim 7, characterized in that: The dynamic mesh optimization unit of the digital twin engine uses non-uniform rational B-splines and NURBS surface models to reconstruct the port terrain. The mesh resolution is dynamically adjusted according to the ship's draft depth and tidal height, with a minimum resolution of 0.5 m x 0.5 m. Incorporating point cloud processing technology, using port point cloud data acquired by drones and underwater robots, the NURBS surface model is updated in real time to improve the accuracy of the digital twin model. The ecological stress analysis unit introduces a physical engine to simulate the impact of the physical environment of port water flow and wind on ships and facilities, achieving more realistic port operation scenario simulation.

9. A port ecological construction evaluation system according to claim 8, characterized in that: The neurofeedback evaluation module of the augmented reality terminal includes: EEG signal acquisition equipment to obtain the decision maker's alpha and beta wave characteristics in real time; The cortical activation map analysis unit uses a convolutional neural network to identify the cognitive matching degree of the solution and output optimization suggestions to the decision optimization layer; At the same time, affective computing technology is introduced to analyze the emotional characteristics of decision makers' facial expressions and voice intonation, and EEG signals are combined to evaluate decision makers' emotional preferences for different construction plans. The virtual scenes generated by the digital twin engine are combined with virtual reality technology to allow decision makers to immersively experience the changes in the port ecological environment under different construction plans, thereby enhancing the intuitiveness and accuracy of decision-making.

10. A port ecological construction evaluation system according to claim 9, characterized in that: The carbon footprint tracking module is used to calculate the carbon balance index of port operations. The specific calculation formula is as follows: ； in: Represents ecological carbon sink, which is calculated based on wetland area and vegetation carbon sequestration efficiency; Represents carbon emissions during the operation phase, which is estimated based on energy consumption and carbon emission factors; represents the embodied carbon emissions during the construction phase. All three are included in the constraints of the multi-objective optimization model; At the same time, the life cycle assessment method is introduced to conduct carbon footprint analysis on the entire life cycle of port construction, operation, maintenance to abandonment, so as to more comprehensively evaluate the carbon emissions of the port; combined with carbon trading market data, a carbon cost assessment model is established to link the carbon balance index with economic costs, providing economic feasibility analysis for the port's ecological construction plan.

Images