Intelligent elevator shaft system self-adaptive to multiple scenes

Abstract

The invention relates to the field of constructional engineering and elevator equipment, and provides an intelligent elevator shaft system self-adaptive to multiple scenes to solve the problems that a traditional elevator shaft building is difficult to adapt, safety monitoring lags behind, and the function is single. According to the system, innovative modules such as self-adaptive flexible connection, improved hybrid algorithm layout optimization and CNN-attention mechanism monitoring are fused, and self-adaptive connection of a well and a building structure, efficient layout optimization, accurate fault monitoring and intelligent energy regulation and control are achieved. And meanwhile, multifunctional equipment is integrated, and full-life-cycle data is managed by using a block chain. The system can remarkably improve the running safety and stability of the elevator, reduce the installation and maintenance cost, bring comfortable and convenient experience to a user and promote intelligent upgrading of a building.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of building engineering and elevator equipment, and particularly relates to an intelligent elevator shaft system self-adapting to multiple scenarios. BACKGROUND

[0002] With the acceleration of urbanization and the development of building technology, as a key equipment for vertical transportation, the elevator shaft exposes a series of problems in actual use, which seriously affects the safety, efficient operation and user experience of the elevator.

[0003] Building adaptation problem: modern building styles are diverse, and structural forms are complex and diverse. Old buildings differ greatly in structure, wall material and space layout due to different construction years. The traditional elevator shaft is relatively fixed in design. When installing an elevator in an old building, it is often difficult to adapt to irregularities in wall structure, inconsistencies in floor height, etc., leading to a difficult installation process, the need for a lot of modification work, increased costs, and possible damage to the original building structure. For newly built special-shaped buildings or buildings with special functions, traditional shafts cannot meet the unique space planning and design requirements of the building, limiting the realization of building functions and aesthetics.

[0004] Safety monitoring and maintenance dilemma: Currently, safety monitoring of elevator shafts mostly relies on regular manual inspection, which is not only inefficient but also greatly influenced by human factors, making it easy to miss inspections. During long-term operation, the shaft structure is affected by various factors such as vibration, impact, corrosion, etc., leading to gradual accumulation of safety hazards such as structural fatigue, loose connections, and worn-out components. Existing monitoring methods cannot monitor the safety status of the shaft in real time and comprehensively, making it difficult to discover and handle these hazards in a timely manner. Once a failure occurs, it may cause the elevator to stop operating or even a safety accident, posing a serious threat to users' life and property safety. In addition, the structural design of traditional shafts is not conducive to maintenance work, with narrow internal space and unreasonable equipment layout, increasing the difficulty and risk of maintenance personnel's operation, resulting in high maintenance costs and low efficiency.

[0005] Single function and lack of intelligence: The existing elevator shaft has a single function, mainly meeting the basic operation needs of the elevator, and lacks integration and expansion of other functions. In the era of rapid development of intelligence, the elevator shaft fails to fully utilize modern technology to achieve intelligent management and collaborative work. For example, the ventilation, lighting, temperature regulation, etc. devices in the shaft often run independently and cannot be intelligently controlled according to the actual operation status of the elevator and user needs, resulting in energy waste. At the same time, the elevator shaft and other intelligent systems of the building (such as security systems, property management systems) lack effective information exchange and linkage mechanisms, and cannot optimize and improve the overall function of the building. SUMMARY

[0006] The present application aims to provide an adaptive multi-element scene intelligent elevator shaft system, aiming to solve many problems faced by existing elevator shafts in practical application, and improve the comprehensive performance of elevator operation and the overall adaptability of the building. It comprises:

[0007] The adaptive flexible connection and variable support module adopts a flexible connection component and an intelligent variable support structure, the angle adjustment range of the flexible connection component is ± 30°, the length adjustment range is 0-400mm, and the support force adjustment range of the intelligent variable support structure is 0-600kN;

[0008] The layout optimization module based on the improved genetic-particle swarm hybrid algorithm evaluates the layout scheme through the fitness function

[0009] Fitness=w1×S+w2×U+w3×C, wherein S represents structural stability, U represents space utilization,

[0010] C represents the complexity of line laying, and w1, w2 and w3 are weight coefficients;

[0011] The safety monitoring and fault diagnosis module based on convolutional neural network-attention mechanism adopts CNN and attention mechanism to process multi-channel sensor data and output safety state evaluation results and fault diagnosis information;

[0012] The intelligent regulation and control and energy management module based on fuzzy logic and reinforcement learning takes multi-source data as the input of the fuzzy logic system and optimizes the regulation and control strategy through reinforcement learning;

[0013] The multifunctional integrated and intelligent interaction module integrates various functional devices and realizes interconnection through the Internet of Things;

[0014] The full life cycle data management module based on blockchain uses blockchain technology to manage elevator shaft full life cycle data.

[0015] Further, the flexible connection component is composed of a high-strength rubber buffer layer and an adjustable metal joint, and the adjustable metal joint is internally provided with a micro motor and a sensor.

[0016] Further, the intelligent variable support structure comprises a hydraulic telescopic rod and a pressure sensor, and the hydraulic telescopic rod automatically adjusts the support force size and direction according to the feedback of the pressure sensor.

[0017] Further, in the improved genetic-particle swarm hybrid algorithm, the particle velocity update formula is

[0018]

[0019] The particle position update formula is

[0020] Further, in the safety monitoring and fault diagnosis module based on convolutional neural network and attention mechanism, the CNN part includes multiple convolutional layers and pooling layers, and the attention mechanism layer generates a weight vector according to the importance of data features.

[0021] Further, in the intelligent regulation and energy management module based on fuzzy logic and reinforcement learning, the fuzzy logic system includes a fuzzification processing unit and a fuzzy rule base, and the reinforcement learning algorithm optimizes the regulation strategy by defining state space, action space and reward function. Further, in the multifunctional integrated and intelligent interaction module, the functional devices include air purification devices, intelligent lighting systems, emergency rescue equipment and environmental monitoring equipment, and the intelligent interaction module provides a user operation interface and a remote monitoring platform for management personnel.

[0023] Further, in the full life cycle data management module based on blockchain, data sharing and access control are realized through smart contracts, and the managed data includes shaft structure parameters, device information, maintenance records, operation data and user feedback.

[0024] Further, the improved genetic-particle swarm hybrid algorithm first uses genetic algorithm for global search, and then uses particle swarm algorithm for local optimization of better solutions.

[0025] Further, in the intelligent regulation and energy management module based on fuzzy logic and reinforcement learning, the reward function is set according to energy utilization efficiency, user comfort and device service life.

[0026] Advantages

[0027] Improve building adaptability: adaptive flexible connection and variable support module can adapt to various building structures and space layouts, whether it is old building elevator installation or special needs of new building, it can realize fast and stable installation, reduce construction cost and damage risk to building structure.

[0028] Enhance safety guarantee: the safety monitoring and fault diagnosis module based on convolutional neural network and attention mechanism can monitor the safety state of the shaft structure in real time and accurately, discover potential faults in time, issue early warning in advance, effectively prevent safety accidents, and protect the safety of passengers' life and property.

[0029] Improve intelligent regulation and energy efficiency: the intelligent regulation and energy management module based on fuzzy logic and reinforcement learning realizes intelligent regulation of elevator operation and shaft equipment, improves energy utilization efficiency, reduces operation cost, and improves user comfort and device service life.

[0030] Multi-functional integration and intelligent interaction: The multi-functional integration and intelligent interaction module enables the elevator shaft to have multiple functions such as air purification, intelligent lighting, emergency rescue, etc., while realizing intelligent interaction between users and the elevator system, making it convenient for users to use and improving management efficiency.

[0031] Optimized data management and collaborative work: The full-life-cycle data management module based on blockchain ensures the safety and traceability of elevator shaft data, promotes the collaborative work between building developers, elevator manufacturers, maintenance companies, regulatory departments, and other stakeholders, and improves the full-life-cycle management level of elevator shafts. BRIEF DESCRIPTION OF DRAWINGS

[0032] Figure 1 Adaptive flexible connection and variable support module schematic diagram;

[0033] Figure 2 Improved genetic-particle swarm hybrid algorithm optimization process schematic diagram;

[0034] Figure 3 Convolutional neural network-attention mechanism model architecture diagram. DETAILED DESCRIPTION

[0035] Example 1

[0036] Through innovative structural design, advanced algorithm model and the application of intelligent technology, the problems of poor building adaptability, difficult safety monitoring and maintenance, single function and lack of intelligence in existing elevator shafts are solved, the safety, stability and intelligence level of elevator operation are improved, the installation and maintenance costs are reduced, and more comfortable and convenient user experience is provided.

[0037] Adaptive flexible connection and variable support module: flexible connection components and intelligent variable support structure are adopted. The flexible connection component is composed of a high-strength rubber buffer layer and an adjustable metal joint. The adjustable metal joint adjusts the angle and length through the built-in micro motor and sensor, the angle adjustment range is ±30°, the length adjustment range is 0-400mm, which can effectively buffer the vibration and impact force during elevator operation, and adapt to the slight deformation of the building structure. The intelligent variable support structure uses hydraulic telescopic rods and pressure sensors to automatically adjust the size and direction of the support force according to the running state of the elevator, the load and the stress change of the building structure, the support force adjustment range is 0-600kN, which ensures the stability of the shaft structure.

[0038] Layout optimization module based on improved genetic-particle swarm hybrid algorithm: The support point layout of the elevator shaft, equipment installation location, and line laying scheme are encoded as chromosomes and particle position vectors. The improved genetic-particle swarm hybrid algorithm combines the global search capability of genetic algorithm and the local search advantage of particle swarm algorithm. By constructing the fitness function Fitness = w1 × S + w2 × U + w3 × C, where S represents the structural stability, U represents the space utilization, C represents the line laying complexity, and w1, w2, w3 are weight coefficients, the pros and cons of the layout scheme are comprehensively evaluated. In the algorithm iteration process, first, the genetic algorithm is used for global search, and the potential optimal solution is found through selection, crossover, and mutation operations; then the particle swarm algorithm is introduced to locally optimize the better solution obtained by the genetic algorithm, and the particles adjust the speed and position according to their own historical optimal position and the historical optimal position of the group, avoiding the algorithm falling into local optimum, and finally the optimal layout scheme is obtained.

[0039] Safety monitoring and fault diagnosis module based on convolutional neural network-attention mechanism: Multiple sensors are installed at key positions of the elevator shaft (such as support beams, connection nodes, guide rails, etc.) to collect real-time data such as vibration, stress, displacement, temperature, etc. A model combining convolutional neural network (CNN) and attention mechanism is used to process the data. CNN is responsible for extracting local features of the data, and attention mechanism automatically assigns weights according to the importance of the data to highlight key information. The model input is multi-channel sensor data, which is convolved and pooled by the CNN layer, then weighted by the attention layer, and finally the safety state evaluation results and fault diagnosis information of the shaft structure are output by the fully connected layer.

[0040] Intelligent control and energy management module based on fuzzy logic and reinforcement learning: The output results of the safety monitoring and fault diagnosis module, elevator operation parameters (such as speed, load, operating floor, etc.), building environment data (such as temperature, humidity, light intensity, etc.), and user demand data (such as user preference settings for temperature and lighting) are used as input variables of the fuzzy logic system. A fuzzy rule base is constructed to generate corresponding control strategies according to different input combinations, such as adjusting the elevator running speed, controlling the working state of ventilation and lighting equipment, adjusting the temperature and humidity in the shaft, etc. At the same time, reinforcement learning algorithm is introduced to optimize the control strategy according to environmental changes and user feedback in actual operation, to improve energy utilization efficiency, enhance user comfort, and ensure stable operation of equipment. Reinforcement learning algorithm defines state space, action space, and reward function, so that the intelligent control module gradually finds the optimal control strategy in the interaction with the environment.

[0041] Multi-functional integration and intelligent interaction module: integrate multiple functional devices in the elevator shaft, such as air purification devices, intelligent lighting systems, emergency rescue equipment, environmental monitoring equipment, etc. These devices are interconnected through Internet of Things technology to form an intelligent interaction network. The intelligent interaction module provides an operation interface for users and management personnel. Users can personalize settings for elevator functions, such as adjusting lighting brightness and selecting air purification modes, through a mobile app or control panel inside the elevator car. Management personnel can remotely monitor the running status of devices in the elevator shaft, perform remote control and management, and achieve fault alarm and rapid response.

[0042] Blockchain-based full life cycle data management module: use blockchain technology to manage the full life cycle data of the elevator shaft from design, manufacturing, installation, operation to scrap. Data includes shaft structure parameters, device information, maintenance records, operation data, user feedback, etc. Through the decentralized and tamper-proof characteristics of blockchain, the authenticity, security and traceability of data are ensured. At the same time, through smart contracts, data sharing and access control are realized. Different stakeholders (such as building developers, elevator manufacturers, maintenance companies, regulatory authorities, etc.) can access and use data according to their permissions, promoting the full life cycle management and collaboration of elevator shafts.

[0043] Innovative connection and support structure design: adaptive flexible connection and variable support modules can automatically adjust the connection and support mode according to the building structure and elevator operation status, effectively solving the poor adaptability problem of traditional shafts, improving the compatibility and stability of the shaft and building structure, and reducing the installation difficulty and impact on the building structure.

[0044] Hybrid algorithm optimization layout: layout optimization module based on improved genetic-particle swarm hybrid algorithm, combining the advantages of both algorithms, realizes efficient optimization of elevator shaft layout, improves structural stability, space utilization, and reduces line laying complexity, providing a guarantee for efficient operation of the elevator system.

[0045] Fusion model safety monitoring and diagnosis: safety monitoring and fault diagnosis module based on convolutional neural network-attention mechanism, combining the characteristics of CNN and attention mechanism, can accurately extract data features, monitor the safety state of the shaft structure in real time and perform fault diagnosis, timely discover potential safety hazards, and ensure the safety of elevator operation.

[0046] Intelligent regulation and energy management strategy: intelligent regulation and energy management module based on fuzzy logic and reinforcement learning, combining the flexible decision-making ability of fuzzy logic and the self-optimization ability of reinforcement learning, realizes intelligent regulation of elevator operation and shaft equipment, improves energy utilization efficiency, enhances user comfort, and ensures stable operation of equipment.

[0047] Multi-functional integration and intelligent interaction: The multi-functional integration and intelligent interaction module realizes the multi-functional integration and intelligent interaction of the elevator shaft, improves the user experience, and facilitates remote monitoring and management by management personnel, improving the management efficiency of the elevator system.

[0048] Blockchain data management and collaboration: The full life cycle data management module based on blockchain uses blockchain technology to ensure the safety and traceability of elevator shaft data, promotes collaboration among stakeholders, and improves the full life cycle management level of the elevator shaft.

[0049] Innovative algorithm model modeling and solving

[0050] Improved genetic-particle swarm hybrid algorithm modeling and solving process:

[0051] Initialization parameters: Set the population size N, crossover probability Pc, mutation probability Pm, and maximum iteration number T of the genetic algorithm, the number of particles n, learning factors c1, c2, and inertia weight w of the particle swarm algorithm. Randomly generate the initial population of the genetic algorithm and the initial particle position and velocity of the particle swarm algorithm. Each individual and particle represents an elevator shaft layout scheme.

[0052] Fitness calculation: Calculate the fitness value of each individual in the genetic algorithm population and each particle in the particle swarm algorithm according to the fitness function Fitness = w1 × S + w2 × U + w3 × C, and evaluate the pros and cons of the layout scheme.

[0053] Genetic algorithm operation: Perform selection operation, select excellent individuals using roulette selection method; Perform crossover operation, perform gene crossover on selected individuals according to crossover probability Pc; Perform mutation operation, mutate individual genes according to mutation probability Pm to generate new population.

[0054] Particle swarm algorithm operation: For particles in the particle swarm algorithm, according to the formula

[0055]

[0056] , the particle position update formula is Update the velocity and position of the particle, where vi,dt is the velocity of particle i in dimension d at the tth iteration, xi,dt is the position of particle i in dimension d at the tth iteration, pi,dt is the historical optimal position of particle i in dimension d, gdt is the historical optimal position of the group in dimension d, r1 and r2 are random numbers between 0 and 1.

[0057] Result fusion and optimization: The better solution obtained by genetic algorithm is used as the initial search range of particle swarm optimization, and particle swarm optimization is used for local search optimization in this range. Compare the results of particle swarm optimization with the optimal solution of genetic algorithm, and keep the better solution.

[0058] Termination condition judgment: when the maximum iteration number T is reached, terminate the algorithm and output the optimal layout scheme; otherwise, return to the fitness calculation step and continue iteration.

[0059] Convolutional neural network-attention mechanism model modeling and solving process:

[0060] Data acquisition and preprocessing: collect multi-channel data such as vibration, stress, displacement and temperature of elevator shaft through sensors, clean the collected data and remove outliers and noise. Then normalize the data to map them to the [0,1] interval to meet the model input requirements. Divide the data according to time sequence to form multiple data samples, each sample containing a certain length of sensor data.

[0061] Model construction: build a convolutional neural network-attention mechanism model, set multiple convolutional and pooling layers in the CNN part. Convolutional layers use different size convolutional kernels to extract local features of data, and pooling layers are used to reduce data dimension and reduce computation. The attention mechanism layer generates a weight vector according to the importance of data features, and weights the features output by CNN. Finally, the features output by the attention mechanism are integrated through the fully connected layer to output the safety state evaluation result and fault diagnosis information of the shaft structure.

[0062] Model training: divide the preprocessed data into training set, validation set and test set. Use cross-entropy loss function as the loss function of the model, and use Adam optimizer to train the model. In the training process, the parameters of the model are constantly adjusted, such as convolution kernel size, step, attention mechanism hyperparameters, etc., to improve the performance of the model. Train the model with the training set and adjust the model parameters with the validation set to prevent overfitting.

[0063] Model evaluation: evaluate the trained model on the test set, and use accuracy, recall, F1 value and other indicators to measure the performance of the model. According to the evaluation results, further optimize and adjust the model until the performance of the model reaches a satisfactory effect.

[0064] Fuzzy logic and reinforcement learning fusion model modeling and solving process:

[0065] Fuzzy logic system construction: Determine the input and output variables of the fuzzy logic system, such as input variables including safety monitoring results, elevator operation parameters, building environment data, user demand data, etc., and output variables including elevator operation speed adjustment, equipment working state control instructions, etc. Fuzzy processing is performed on the input variables, and fuzzy subsets and membership functions are defined. According to the experience of field experts and actual operation data, a fuzzy rule base is constructed, for example, "if the shaft vibration is large and the elevator load is close to the rated load, then reduce the elevator running speed and increase the power of the ventilation equipment."

[0066] Reinforcement learning setup: Define the state space of reinforcement learning, including the input variables of the fuzzy logic system and part of the internal state of the elevator system; define the action space, i.e. the possible values of the output variables of the fuzzy logic system; design a reward function according to energy utilization efficiency, user comfort, equipment service life, etc. For example, if energy consumption is reduced, user comfort is improved, and equipment service life is extended, positive rewards are given, otherwise negative rewards are given.

[0067] Model training and optimization: In actual operation, the fuzzy logic system generates an initial control strategy according to the input variables, and the reinforcement learning algorithm obtains reward feedback according to the current state and action. By continuously adjusting the parameters of the fuzzy logic system (such as the weights of fuzzy rules) and the parameters of the reinforcement learning algorithm (such as learning rate, discount factor), the control strategy is optimized to maximize the reward of the system in the long run.

[0068] Adaptive flexible connection and variable support module implementation: In the elevator installation project of an old community, the building walls are masonry structures and there is a certain settlement. When installing the elevator shaft of the invention, the flexible connection components of the adaptive flexible connection and variable support module automatically adjust the angle and length to compensate for the differences caused by wall settlement, and the intelligent variable support structure adjusts the support force in real time according to the stress condition of the wall to ensure the stability of the shaft structure after installation. After long-term monitoring, the vibration and noise of the shaft during elevator operation are significantly reduced, and the building structure does not appear abnormal stress condition.

[0069] Improved genetic-particle swarm hybrid algorithm layout optimization implementation: In the elevator shaft design of a newly built commercial complex, an improved genetic-particle swarm hybrid algorithm is used for layout optimization. The population size of genetic algorithm is set to 50, the crossover probability is 0.8, the mutation probability is 0.2, and the maximum iteration number is 200; the particle number of particle swarm algorithm is 30, the learning factor c1=c2=1.5, and the inertia weight w=0.8. After algorithm optimization, the structural stability of the shaft is improved by 30%, the space utilization rate is improved by 25%, the line laying complexity is reduced by 20%, and the interference between equipment is reduced, and the operation efficiency is significantly improved.

Claims

1. An adaptive intelligent elevator shaft system for multiple scenarios, characterized in that, include: The adaptive flexible connection and variable support module adopts a flexible connection component and an intelligent variable support structure. The angle adjustment range of the flexible connection component is ±30°, the length adjustment range is 0-400mm, and the support force adjustment range of the intelligent variable support structure is 0-600kN. The layout optimization module based on the improved genetic-particle swarm optimization algorithm uses a fitness function. Fitness = w1×S + w2×U + w3×C to evaluate the layout scheme, where S represents structural stability, U represents space utilization, C represents the complexity of line laying, and w1, w2, and w3 are weighting coefficients. The safety monitoring and fault diagnosis module based on convolutional neural network-attention mechanism uses CNN and attention mechanism to process multi-channel sensor data and output safety status assessment results and fault diagnosis information. The intelligent regulation and energy management module based on fuzzy logic and reinforcement learning uses multi-source data as input to the fuzzy logic system and optimizes the regulation strategy through reinforcement learning. A multi-functional integrated and intelligent interaction module that integrates multiple functional devices and achieves interconnection through the Internet of Things; The blockchain-based full lifecycle data management module utilizes blockchain technology to manage elevator shaft data throughout its entire lifecycle.

2. The adaptive multi-scenario intelligent elevator shaft system according to claim 1, characterized in that, The flexible connection assembly consists of a high-strength rubber buffer layer and an adjustable metal joint, which incorporates a micro motor and a sensor.

3. The adaptive multi-scenario intelligent elevator shaft system according to claim 1, characterized in that, The intelligent variable support structure includes a hydraulic telescopic rod and a pressure sensor. The hydraulic telescopic rod automatically adjusts the magnitude and direction of the support force based on feedback from the pressure sensor.

4. The adaptive multi-scenario intelligent elevator shaft system according to claim 1, characterized in that, In the improved genetic-particle swarm optimization algorithm, the particle velocity update formula is as follows: The particle position update formula is:

5. The adaptive multi-scenario intelligent elevator shaft system according to claim 1, characterized in that, In the security monitoring and fault diagnosis module based on convolutional neural network-attention mechanism, the CNN part includes multiple convolutional layers and pooling layers, and the attention mechanism layer generates weight vectors according to the importance of data features.

6. The adaptive multi-scenario intelligent elevator shaft system according to claim 1, characterized in that, In the intelligent regulation and energy management module based on fuzzy logic and reinforcement learning, the fuzzy logic system includes a fuzzification processing unit and a fuzzy rule base, and the reinforcement learning algorithm optimizes the regulation strategy by defining the state space, action space and reward function.

7. The adaptive multi-scenario intelligent elevator shaft system according to claim 1, characterized in that, The multifunctional integrated and intelligent interaction module includes functional devices such as air purification devices, intelligent lighting systems, emergency rescue equipment, and environmental monitoring equipment. The intelligent interaction module provides a user interface and a remote monitoring platform for management personnel.

8. The adaptive multi-scenario intelligent elevator shaft system according to claim 1, characterized in that, The blockchain-based full lifecycle data management module uses smart contracts to achieve data sharing and access control. The managed data includes shaft structure parameters, equipment information, maintenance records, operating data, and user feedback.

9. The adaptive multi-scenario intelligent elevator shaft system according to claim 1, characterized in that, The improved genetic-particle swarm optimization hybrid algorithm first uses a genetic algorithm for global search, and then uses a particle swarm optimization algorithm to perform local optimization on the better solution.

10. The adaptive multi-scenario intelligent elevator shaft system according to claim 1, characterized in that, In the intelligent regulation and energy management module based on fuzzy logic and reinforcement learning, the reward function is set according to energy utilization efficiency, user comfort, and equipment lifespan.

Images