Intelligent construction hoist control system

Abstract

The invention discloses an intelligent construction hoist control system, which performs data analysis and feature fusion on signal strength, bit error rate and equipment state in a communication process by adopting a data processing technology based on deep learning so as to obtain communication constraint condition features fused with multi-modal information. And semantic guidance interaction is carried out on the communication constraint condition and the communication demand to fully understand potential association between the communication demand and the constraint, so that intelligent recommendation optimization of the SF value is realized, optimal configuration of communication parameters of the LoRa module is effectively realized, the anti-interference capability and transmission reliability of the signal are enhanced, and the communication performance of the LoRa module is improved. The communication requirements of different areas can be met, and therefore it is ensured that the elevator control system can keep efficient and stable communication under various conditions.

Description

Technical Field

[0001] The present application relates to the field of intelligent control, and more specifically, to an intelligent construction hoist control system. Background Art

[0002] With the acceleration of urbanization, there are more and more high-rise buildings. During the construction of high-rise buildings, elevators are needed to transport personnel and materials. Traditional elevator control systems mostly use wired connections for signal transmission, which not only increases the difficulty and cost of wiring, but also limits the flexibility and scalability of the elevator system.

[0003] In recent years, with the rapid development of Internet of Things (IoT) technology, wireless communication technology has been widely used in elevator systems. At present, there are many elevator control systems based on wireless communication technology on the market, most of which use wireless communication technologies such as Zigbee, Wi-Fi or Bluetooth to achieve communication between various parts of the elevator by building a local network. However, these technologies have certain limitations in practical applications, such as short communication distance, high power consumption, and weak anti-interference ability. Especially in high-rise buildings or large projects, the limitation of signal coverage may cause some floors or areas to be unable to effectively receive control instructions.

[0004] LoRa (Long Range Radio) technology provides a new solution for the intelligent and wireless elevator system with its long-distance communication, low power consumption, and low cost. However, although the LoRa module has the advantages of long-distance communication and low power consumption, its transmission rate is relatively low and may be interfered in complex electromagnetic environments, resulting in reduced communication stability.

[0005] Therefore, an optimized intelligent construction hoist control system is expected to dynamically adjust the communication parameters of the LoRa model according to the specific conditions of different areas to ensure the reliability and efficiency of communication. Summary of the invention

[0006] In order to solve the above technical problems, this application is proposed. The embodiment of this application provides an intelligent construction elevator control system, which uses deep learning-based data processing technology to perform data analysis and feature fusion on the signal strength, bit error rate and equipment status in the communication process to obtain the communication constraint condition characteristics that fuse multimodal information, and then through semantically guided interaction between communication constraints and communication requirements, to fully understand the potential relationship between communication requirements and constraints, so as to achieve intelligent recommendation optimization of SF values, effectively realize the optimal configuration of LoRa module communication parameters, enhance the signal's anti-interference ability and transmission reliability, so that it can adapt to the communication needs of different regions, thereby ensuring that the elevator control system can maintain efficient and stable communication under various conditions.

[0007] The present application provides an intelligent construction hoist control system, which includes: The external call box is used to receive the user's call command and transmit the call command to the elevator controller through the wireless module; A key box installed in the elevator car, used to receive control commands from the user; The elevator controller is used to process instructions from the hall call box and the key box to control the speed and direction of the elevator; The elevator controller is further used to dynamically adjust the communication parameters of the LoRa module based on the communication condition information of the area, wherein the communication condition information includes a time queue of a signal strength value, a time queue of a bit error rate, a natural language description of communication requirements, and a device status; Wherein, the elevator controller comprises: A data vectorization module, used for vectorizing the time queue of the signal strength value, the time queue of the bit error rate, the natural language description of the communication demand and the device state to obtain a signal strength time series associated implicit feature vector, a bit error rate time series associated implicit feature vector, a communication demand semantic coding feature vector and a device state semantic embedded coding vector; A communication constraint multimodal information fusion module, used for jointly encoding the signal strength time series associated implicit feature vector, the bit error rate time series associated implicit feature vector and the device state semantic embedded coding vector to obtain a communication constraint condition multimodal fusion representation vector; The feature guided interaction module is used to perform feature guided interaction based on implicit association features on the communication constraint multimodal fusion representation vector and the communication requirement semantic encoding feature vector to obtain a communication constraint-communication requirement significant interaction fusion representation vector.

[0008] Specifically, the elevator controller further includes: The communication parameter optimization module is used to generate an optimization instruction for the communication parameters based on the communication constraint condition-communication demand significant interaction fusion representation vector, wherein the optimization instruction includes a recommended SF value.

[0009] Specifically, the data vectorization module includes: A time series encoding unit, used for performing time series encoding on the time series of the signal strength value and the time series of the bit error rate respectively using an LSTM model to obtain an implicit feature vector associated with the signal strength time series and an implicit feature vector associated with the bit error rate time series; An embedding coding unit, used for embedding the device state using a device state embedding coding matrix to obtain a device state semantic embedding coding vector; The semantic encoding unit is used to perform semantic embedding encoding on the natural language description of the communication demand using the Bert model to obtain the semantic encoding feature vector of the communication demand.

[0010] Specifically, the communication constraint multimodal information fusion module is used to: The signal strength time series associated implicit feature vector, the bit error rate time series associated implicit feature vector and the device state semantic embedding coding vector are input into a multimodal joint encoder based on a fully connected layer to obtain the communication constraint condition multimodal fusion representation vector.

[0011] Specifically, the feature-guided interaction module includes: A condition information construction unit, used to extract the implicit correlation features between the communication constraint condition multimodal fusion representation vector and the communication requirement semantic encoding feature vector as a condition feature vector; A feature modulation unit, configured to perform feature modulation on the communication constraint condition multimodal fusion representation vector and the communication requirement semantic coding feature vector based on the semantic relevance of the communication constraint condition multimodal fusion representation vector and the communication requirement semantic coding feature vector relative to the condition feature vector to obtain a modulated communication constraint condition multimodal fusion representation vector and a modulated communication requirement semantic coding feature vector; A cross-domain interaction coding unit is used to perform cross-domain interaction coding based on a converter structure on the multimodal fusion representation vector of the modulated communication constraint condition, the semantic coding feature vector of the modulated communication requirement, and the conditional feature vector to obtain the communication constraint condition-communication requirement significant interaction fusion representation vector.

[0012] Specifically, the condition information construction unit includes: An implicit correlation feature extraction subunit, used for inputting the communication constraint condition multimodal fusion representation vector and the communication requirement semantic encoding feature vector into an implicit correlation feature capture network to obtain a communication constraint condition-communication requirement implicit correlation feature vector; The feature activation subunit is used to perform feature activation based on the Sigmoid function on the communication constraint condition-communication requirement implicit association feature vector to obtain the conditional feature vector.

[0013] Specifically, the characteristic modulation unit includes: A first semantic relevance measurement subunit, used to calculate a first semantic relevance factor of the communication constraint condition multimodal fusion representation vector relative to the condition feature vector; A second semantic relevance measurement subunit, used to calculate a second semantic relevance factor of the communication requirement semantic encoding feature vector with respect to the condition feature vector; A weighted modulation subunit is used to normalize the first semantic relevance factor and the second semantic relevance factor, and use the normalized first semantic relevance factor and the second semantic relevance factor to weighted modulate the communication constraint multimodal fusion representation vector and the communication requirement semantic coding feature vector to obtain the modulated communication constraint multimodal fusion representation vector and the modulated communication requirement semantic coding feature vector.

[0014] Specifically, the cross-domain interaction encoding unit is used to: Taking the multimodal fusion representation vector of the post-modulation communication constraint as the query vector, the semantically encoded feature vector of the post-modulation communication requirement as the key vector and the conditional feature vector as the value vector, the multimodal fusion representation vector of the post-modulation communication constraint, the semantically encoded feature vector of the post-modulation communication requirement and the conditional feature vector are input into a feature-to-feature significant guided interaction module based on a converter structure to obtain the communication constraint-communication requirement significant interaction fusion representation vector.

[0015] Specifically, the communication parameter optimization module is used to: The communication constraint condition-communication demand significant interaction fusion representation vector is input into a decoder-based communication parameter optimization controller to obtain the optimization instruction.

[0016] Specifically, it also includes a touch screen interface, which is integrated in the key box and is used to provide elevator status information.

[0017] This application has at least the following technical effects: Compared with the prior art, the present application provides an intelligent construction elevator control system, which uses deep learning-based data processing technology to perform data analysis and feature fusion on the signal strength, bit error rate and equipment status in the communication process, so as to obtain the communication constraint condition characteristics that integrate multimodal information, and then semantically guide the interaction between the communication constraints and the communication requirements to fully understand the potential relationship between the communication requirements and the constraints, thereby realizing intelligent recommendation optimization of the SF value, effectively realizing the optimal configuration of the LoRa module communication parameters, enhancing the signal's anti-interference ability and transmission reliability, so that it can adapt to the communication needs of different areas, thereby ensuring that the elevator control system can maintain efficient and stable communication under various conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0019] Figure 1 is a block diagram of an intelligent construction elevator control system according to an embodiment of the present application; Figure 2 A data flow diagram of a smart construction elevator control system according to an embodiment of the present application; Figure 3 It is a block diagram of an elevator controller in an intelligent construction elevator control system according to an embodiment of the present application. DETAILED DESCRIPTION

[0020] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0021] As shown in this application and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0022] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.

[0023] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps may be processed in reverse order or simultaneously as required. Meanwhile, other operations may also be added to these processes, or a certain step or several steps of operations may be removed from these processes.

[0024] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0025] In the technical solution of the present application, an intelligent construction hoist control system is proposed. Figure 1 4 is a block diagram of an intelligent construction elevator control system according to an embodiment of the present application. Figure 2 Schematic diagram of data flow of the intelligent construction elevator control system according to an embodiment of the present application. Figure 1 and Figure 2 As shown, according to an embodiment of the present application, the intelligent construction elevator control system 300 includes: an external call box 310, which is used to accept the user's call instructions and transmit the call instructions to the elevator controller through a wireless module; a key box 320 installed in the elevator car, which is used to accept the user's control instructions; the elevator controller 330, which is used to process instructions from the external call box and the key box to control the speed and direction of the elevator; a touch screen interface 340, which is integrated in the key box and is used to provide elevator status information.

[0026] In particular, the external call box 310 is used to receive the user's call instructions and transmit the call instructions to the elevator controller through the wireless module. Among them, the external call box is mainly responsible for receiving the user's call instructions from outside the elevator. These instructions usually include the floor the user wants to reach and the choice of going up or down. The external call box is one of the very important components in the elevator system. It not only improves the user experience, but also enables the elevator system to operate more efficiently.

[0027] In particular, the key box 320 installed in the elevator car is used to receive control instructions from the user, including selecting a target floor. In addition, the key box is usually equipped with safety facilities such as an emergency stop button. In case of an emergency, passengers can press the button to immediately stop the elevator operation to ensure the safety of the passengers.

[0028] In particular, the elevator controller 330 is used to process instructions from the outside call box and the key box to control the speed and direction of the elevator. Among them, the wireless module in the outside call box adopts LoRa communication technology, which can realize long-distance wireless communication. The outside call box realizes communication and configuration with the main control system through a series of parameter settings, including but not limited to building number, current floor setting, host slave setting, learning extension number, number of learned extensions, extension deletion function, floor bias, LoRa module communication parameters and CAT1 module signal strength. By pressing a specific key combination, the user can enter the parameter setting interface and set and modify parameters through keys such as "plus", "minus" and "confirm". In order to ensure stable communication between the outside call box and the main control system, the LoRa module in the outside call box needs to set communication parameters that match the main control system, such as LoRa module channel, target address, rate level and forward error correction parameters. By setting the communication parameters to be consistent with the main control system, it is ensured that the outside call box can accurately receive and send control instructions. In LoRa communication, the spreading factor (SF) is an important parameter in LoRa technology, which determines the number of symbols sent for each information bit and reflects the degree of information spreading. In LoRa communication, the smaller the SF value, the higher the transmission rate but the smaller the coverage range; and vice versa. Therefore, when applying LoRa technology, it is usually necessary to select an appropriate SF value based on the actual required distance and real-time requirements. For example, in application scenarios that require long-distance transmission and are not sensitive to delays, a larger SF value can be selected. Based on this, in the technical solution of the present application, in order to balance the transmission distance and transmission rate, the elevator controller is further used to dynamically adjust the communication parameters of the LoRa module based on the communication condition information of the area, wherein the communication condition information includes a time queue of signal strength values, a time queue of bit error rates, a natural language description of communication requirements, and device status. In a specific example of the present application, such as Figure 3As shown, the elevator controller 330 includes: a data vectorization module 331, which is used to vectorize the time queue of the signal strength value, the time queue of the bit error rate, the natural language description of the communication demand and the device state to obtain the signal strength time series associated implicit feature vector, the bit error rate time series associated implicit feature vector, the communication demand semantic coding feature vector and the device state semantic embedded coding vector; a communication constraint multimodal information fusion module 332, which is used to jointly encode the signal strength time series associated implicit feature vector, the bit error rate time series associated implicit feature vector and the device state semantic embedded coding vector to obtain a communication constraint condition multimodal fusion representation vector; a feature guided interaction module 333, which is used to perform feature guided interaction based on implicit association features on the communication constraint condition multimodal fusion representation vector and the communication demand semantic coding feature vector to obtain a communication constraint condition-communication demand significant interaction fusion representation vector; a communication parameter optimization module 334, which is used to generate an optimization instruction for communication parameters based on the communication constraint condition-communication demand significant interaction fusion representation vector, and the optimization instruction includes a recommended SF value.

[0029] Specifically, the data vectorization module 331 is used to vectorize the time queue of the signal strength value, the time queue of the bit error rate, the natural language description of the communication demand and the device state to obtain the signal strength time series associated implicit feature vector, the bit error rate time series associated implicit feature vector, the communication demand semantic coding feature vector and the device state semantic embedded coding vector. Among them, the signal strength refers to the power level of the received radio signal, which directly affects the quality and distance of data transmission. The bit error rate (Bit Error Rate, BER) represents the proportion of erroneous bits in the bits transmitted within a certain period of time, and is an important parameter for measuring the quality of the communication link. A low bit error rate indicates that there are fewer errors in the data transmission process and the communication quality is good; while a high bit error rate may indicate that the signal is severely interfered or attenuated, resulting in a decrease in the quality of data transmission. The device state provides real-time feedback on the operation and communication system of the elevator, including but not limited to information such as equipment operation mode, temperature, voltage, current, fault code, etc. In the technical solution of the present application, real-time monitoring and comprehensive analysis of signal strength, bit error rate and device state help to achieve a comprehensive understanding of the communication environment. In addition, the natural language description of the communication requirements is the user's specific requirements and expectations for the operation of the elevator, such as "ensuring that the transmission delay of the control command does not exceed 5 seconds", "maintaining the bit error rate below 1%", etc. By combining the user's communication requirements and the actual communication environment, it helps to more accurately guide the communication behavior of the LoRa module.

[0030] Considering that during the communication process, the signal strength and bit error rate change over time and have strong time dependence. Therefore, in order to effectively capture the long-term dependencies in the time series data, so as to better understand the changing trends of the signal strength and bit error rate, this application uses the LSTM model to perform time series encoding on the time queue of the signal strength value and the time queue of the bit error rate, respectively, so as to utilize the long-term memory characteristics of the LSTM model to capture the changing patterns and trends of the signal strength and bit error rate in the global time domain, generate the signal strength time series associated implicit feature vector and the bit error rate time series associated implicit feature vector, so as to characterize the time series dynamic characteristics of the signal strength and the bit error rate. Among them, the Long Short-Term Memory (LSTM) network is a special recurrent neural network (RNN) that controls the flow of information by introducing three gating mechanisms (input gate, forget gate and output gate), thereby alleviating the gradient vanishing and gradient exploding problems.

[0031] Next, since the device state usually contains multiple different types of data, in order to further unify the data representation and achieve a global understanding of the device state, so as to facilitate subsequent data fusion analysis, the present application uses a device state embedding coding matrix to embed the device state, so as to map the multi-dimensional discrete data of the device state into a continuous vector space, so as to achieve global semantic information integration and low-dimensional feature representation of the device state, thereby obtaining a device state semantic embedding coding vector. In one embodiment of the present application, the device state embedding coding matrix is ​​obtained by pre-training the Word2Vec model. Among them, Word2Vec is a shallow neural network model for generating word embeddings. Word2Vec maps vocabulary to a high-dimensional vector space by learning a large amount of text data, so that semantically similar words have similar positions in the vector space.

[0032] At the same time, considering that the communication demand is a natural language description with rich contextual semantic information. Therefore, in order to achieve semantic understanding of communication needs, this application uses the Bert model to perform semantic embedding encoding on the natural language description of the communication needs, so as to utilize the bidirectional semantic encoding capability of the Bert model, while considering the forward and backward contextual information of each vocabulary unit in the natural language description of the communication needs, so as to more accurately capture the semantic associations and contextual information therein, and generate a semantic embedding encoding vector for communication needs. Among them, BERT (Bidirectional Encoder Representations from Transformers) is a pre-trained language model. Through the deep bidirectional transformer (Transformer) architecture, BERT can effectively capture the contextual information in the text and generate high-quality word vectors and sentence representations.

[0033] Specifically, the communication constraint multimodal information fusion module 332 is used to jointly encode the signal strength time series associated implicit feature vector, the bit error rate time series associated implicit feature vector and the device state semantic embedded coding vector to obtain a communication constraint condition multimodal fusion representation vector. It should be understood that since the signal strength, bit error rate and device state respectively reflect the quality of the communication link, the reliability of data transmission and the working state of the device, the three together describe the comprehensive status of the communication environment and constitute the constraints that affect the communication quality. Therefore, in order to more comprehensively understand the current communication environment, the present application further uses a multimodal joint encoder based on a fully connected layer to jointly encode the signal strength time series associated implicit feature vector, the bit error rate time series associated implicit feature vector and the device state semantic embedded coding vector, so as to comprehensively consider the multidimensional characteristics of the communication environment, and capture the interaction and influence between the communication constraint features of different dimensions, thereby generating a communication constraint condition multimodal fusion representation vector.

[0034] Specifically, the feature guided interaction module 333 is used to perform feature-guided interaction based on implicit association features on the communication constraint multimodal fusion representation vector and the communication requirement semantic encoding feature vector to obtain a communication constraint-communication requirement significant interaction fusion representation vector. It should be understood that the communication constraints reflect the actual situation of the current communication environment, while the communication requirements express the expectations for communication performance. In the process of optimizing the configuration of communication parameters, in order to ensure that the communication constraints and communication requirements can be effectively matched, the present application further proposes a feature-guided interaction method based on implicit association features to perform information interaction processing on the communication constraint multimodal fusion representation vector and the communication requirement semantic encoding feature vector, so as to identify and strengthen the important association patterns between the communication constraints and the communication requirements, thereby achieving deep interaction between the two.

[0035] Specifically, first, the implicit correlation features between the communication constraint multimodal fusion representation vector and the communication requirement semantic encoding feature vector are extracted as conditional feature vectors. The specific steps include: inputting the communication constraint multimodal fusion representation vector and the communication requirement semantic encoding feature vector into an implicit correlation feature capture network to obtain a communication constraint-communication requirement implicit correlation feature vector; performing feature activation based on the Sigmoid function on the communication constraint-communication requirement implicit correlation feature vector to obtain the conditional feature vector. That is, the implicit correlation features shared between the communication constraint multimodal fusion representation vector and the communication requirement semantic encoding feature vector are mined through an implicit correlation feature capture network, and the extracted implicit correlation features are nonlinearly transformed using the Sigmoid function to emphasize key information and suppress irrelevant noise, thereby constructing a conditional feature vector to guide the subsequent feature interaction fusion process.

[0036] More specifically, the communication constraint multimodal fusion representation vector and the communication requirement semantic encoding feature vector are input into an implicit association feature capture network to obtain a communication constraint-communication requirement implicit association feature vector, including: fusing the communication constraint multimodal fusion representation vector and the communication requirement semantic encoding feature vector at position points and then passing them through a neural network layer based on a tanh function to obtain the communication constraint-communication requirement implicit association feature vector.

[0037] Next, based on the semantic relevance of the communication constraint multimodal fusion representation vector and the communication requirement semantic coding feature vector relative to the condition feature vector, feature modulation is performed on the communication constraint multimodal fusion representation vector and the communication requirement semantic coding feature vector to obtain the modulated communication constraint multimodal fusion representation vector and the modulated communication requirement semantic coding feature vector. The specific steps include: calculating the first semantic relevance factor of the communication constraint multimodal fusion representation vector relative to the condition feature vector; calculating the second semantic relevance factor of the communication requirement semantic coding feature vector to the condition feature vector; normalizing the first semantic relevance factor and the second semantic relevance factor, and using the normalized first semantic relevance factor and the second semantic relevance factor to perform weighted modulation on the communication constraint multimodal fusion representation vector and the communication requirement semantic coding feature vector to obtain the modulated communication constraint multimodal fusion representation vector and the modulated communication requirement semantic coding feature vector. That is, the communication constraint multimodal fusion representation vector and the communication requirement semantic coding feature vector are respectively evaluated for correlation with the conditional feature vector to determine the importance of the two for the associated feature representation, and thereby weighted modulate the original communication constraint multimodal fusion representation vector and the communication requirement semantic coding feature vector, thereby effectively emphasizing the key information in the two and suppressing irrelevant information.

[0038] More specifically, calculating the first semantic relevance factor of the communication constraint multimodal fusion representation vector relative to the conditional feature vector includes: calculating the dot division vector between the communication constraint multimodal fusion representation vector and the conditional feature vector, and calculating the base-two logarithmic value of the absolute value of each eigenvalue in the dot division vector to obtain the communication constraint semantic relevance representation vector; calculating the dot product vector between the communication constraint semantic relevance representation vector and the communication constraint multimodal fusion representation vector, and calculating an exponential function value with base e and the sum of the eigenvalues ​​of the dot product vector as the exponent to obtain the first semantic relevance factor.

[0039] Finally, the multimodal fusion representation vector of the post-modulation communication constraint, the semantic coding feature vector of the post-modulation communication requirement and the conditional feature vector are cross-domain interactively encoded based on the converter structure to obtain the communication constraint-communication requirement significant interactive fusion representation vector. In a specific example of the present application, the post-modulation communication constraint condition multimodal fusion representation vector is used as a query vector, the post-modulation communication requirement semantic coding feature vector is used as a key vector, and the conditional feature vector is used as a value vector. The post-modulation communication constraint condition multimodal fusion representation vector, the post-modulation communication requirement semantic coding feature vector, and the conditional feature vector are input into a feature-based significant guided interaction module based on a converter structure to obtain the communication constraint condition-communication requirement significant interaction fusion representation vector. The specific steps include: multiplying the post-modulation communication constraint condition multimodal fusion representation vector by the transposed vector of the post-modulation communication requirement semantic coding feature vector and then dividing it by the square root of the length of the communication requirement semantic coding feature vector to obtain the communication constraint condition-communication requirement attention score matrix; multiplying the communication constraint condition-communication requirement attention score matrix by the conditional feature vector after passing the softmax function to obtain the communication constraint condition-communication requirement significant interaction fusion representation vector. In this way, deep interactive fusion of cross-modal features is achieved, and a communication constraint condition-communication requirement significant interaction fusion representation vector is generated, thereby achieving a comprehensive understanding of the communication environment and requirements.

[0040] In summary, in the above embodiment, the communication constraint multimodal fusion representation vector and the communication requirement semantic encoding feature vector are subjected to feature-guided interaction based on implicit association features to obtain a communication constraint-communication requirement significant interaction fusion representation vector, including: processing the communication constraint multimodal fusion representation vector and the communication requirement semantic encoding feature vector with the following feature interaction fusion formula to obtain the communication constraint-communication requirement significant interaction fusion representation vector, wherein the feature interaction fusion formula is: in, represents the multimodal fusion representation vector of the communication constraint condition, represents the communication requirement semantic encoding feature vector, Indicates adding by position point. represents the implicitly associated feature vector of communication constraints and communication requirements, and Represent the weight matrix and bias parameters of the neural network layer, respectively. represents the hyperbolic tangent function, represents the sigmoid activation function, represents the conditional eigenvector, , and Respectively represent the first eigenvalue, the first eigenvalue in the communication requirement semantic encoding feature vector eigenvalue and the conditional eigenvector eigenvalues, is the absolute value symbol, represents the logarithmic function with base 2, represents the exponential function with base e, represents the length of the multimodal fusion representation vector of the communication constraint condition, and the lengths of the multimodal fusion representation vector of the communication constraint condition, the communication requirement semantic encoding feature vector and the condition feature vector are the same, and denote the first semantic relevance factor and the second semantic relevance factor respectively, and denote the normalized first semantic relevance factor and the second semantic relevance factor, respectively. and They represent the multimodal fusion representation vector of the post-modulation communication constraint conditions and the semantic encoding feature vector of the post-modulation communication requirement, respectively. represents the transpose of a vector, represents the matrix multiplication operation, is the normalized exponential function, A significant interaction fusion representation vector is represented by representing the communication constraint-communication requirement.

[0041] Specifically, the communication parameter optimization module 334 is used to generate an optimization instruction for the communication parameters based on the communication constraint condition-communication requirement significant interaction fusion representation vector, and the optimization instruction includes a recommended SF value. In a specific example of the present application, the communication constraint condition-communication requirement significant interaction fusion representation vector is input into a decoder-based communication parameter optimization controller to perform decoding optimization of the communication parameters, so as to realize intelligent recommendation of the SF value of the LoRa module, thereby maximizing the satisfaction of the user's communication needs under the conditional constraints of the communication environment.

[0042] In the technical solution of the present application, the communication constraint multimodal fusion representation vector and the communication requirement semantic coding feature vector respectively represent the communication constraint multimodal fusion coding features and the text semantic coding features of the communication requirements, which are composed of the timing features of signal strength, bit error rate and device status. In this way, when it is input into the inter-feature significant guided interaction module with the joint implicit feature as the conditional information, considering the modal differences and feature coding differences between the communication constraint multimodal fusion representation vector and the communication requirement semantic coding feature vector, this will make the extraction of the joint implicit feature insufficient and inaccurate, resulting in the feature manifold of the communication constraint-communication requirement significant interaction fusion representation vector obtained by the inter-feature significant guided interaction with the joint implicit feature as the conditional information. There is expression sparsity and uneven distribution in the high-dimensional feature space. Therefore, it is expected to improve the semantically consistent aggregation expression effect of the communication constraint-communication requirement significant interaction fusion representation vector obtained by the inter-feature significant guided interaction.

[0043] Based on this, in a preferred embodiment, the communication constraint condition-communication demand significant interaction fusion representation vector is input into a decoder-based communication parameter optimization controller to obtain an optimization instruction, including: The feature set of the communication constraint condition-communication requirement significant interaction fusion representation vector is clustered to obtain a feature set within the communication constraint condition-communication requirement significant interaction fusion class and a feature set outside the communication constraint condition-communication requirement significant interaction fusion class, that is: in, It represents the communication constraint-communication requirement significant interaction fusion intra-class feature set, The first feature set of the communication constraint-communication requirement significant interaction fusion class eigenvalues, The first one represents the communication constraint-communication requirement significant interaction fusion out-of-class feature set eigenvalues; The ratio of the number of eigenvalues ​​in the feature set of the communication constraint condition-communication requirement significant interaction fusion class to the number of eigenvalues ​​in the feature set of the communication constraint condition-communication requirement significant interaction fusion representation vector is calculated, that is: in, represents the number of eigenvalues ​​in the feature set of the communication constraint-communication demand significant interaction fusion class, The number of eigenvalues ​​of the feature set representing the communication constraint-communication demand significant interaction fusion representation vector, It indicates the ratio; Calculate the absolute value sum of all feature values ​​in the feature set of the communication constraint-communication requirement significant interaction fusion class The power of the communication constraint condition-communication requirement significantly interacts with the sum of the absolute values ​​of all eigenvalues ​​in the feature set of the representation vector The ratio of the powers is used to obtain the communication constraint condition-communication demand significant interaction fusion modulation weight, that is: in, Each eigenvalue in the feature set representing the communication constraint-communication requirement significant interaction fusion representation vector, It represents the communication constraint-communication demand significant interaction fusion modulation weight; Calculate the sum of squares of all feature values ​​in the feature set of the communication constraint-communication requirement significant interaction fusion class The power significantly interacts with the communication constraint - the communication requirement, and the sum of the squares of all eigenvalues ​​in the feature set of the fusion representation vector The ratio of the powers is used to obtain the communication constraint condition-communication demand significant interaction fusion harmonic weight, that is: in, It represents the communication constraint condition-communication demand significant interaction fusion and reconciliation weight; For each feature value in the feature set within the communication constraint condition-communication requirement significant interaction fusion class, calculate the product of the feature value and the communication constraint condition-communication requirement significant interaction fusion harmonic weight value and add the communication constraint condition-communication requirement significant interaction fusion modulation weight value to obtain the optimized feature value, that is: in, represents the first intra-class feature set based on the communication constraint condition-communication requirement significant interaction fusion Optimized eigenvalues; For each feature value in the communication constraint condition-communication requirement significant interaction fusion out-of-class feature set, the product of the feature value and the communication constraint condition-communication requirement significant interaction fusion adjustment weight is calculated to obtain the optimized feature value, that is: in, represents the first out-of-class feature set based on the communication constraint condition-communication requirement significant interaction fusion Optimized eigenvalues; The optimized communication constraint - communication demand significant interaction fusion representation vector based on the optimized feature values ​​of the communication constraint - communication demand significant interaction fusion class feature set and the communication constraint - communication demand significant interaction fusion class feature set is passed through a decoder-based communication parameter optimization controller to obtain the optimization instruction.

[0044] While performing feature clustering on the communication constraint condition-communication requirement significant interaction fusion representation vector, the interaction description of the key feature information of the communication constraint condition-communication requirement significant interaction fusion representation vector in the clustering process is carried out, and the geometric equivariant topology of the feature is constructed by low-rank harmonic modulation of the clustering features of the communication constraint condition-communication requirement significant interaction fusion representation vector and the equivariance of the feature as a whole, so as to obtain the graphical distribution translation and rotation symmetry of the clustering features of the communication constraint condition-communication requirement significant interaction fusion representation vector relative to the feature as a whole, so as to introduce geometric message passing into the feature expression of the communication constraint condition-communication requirement significant interaction fusion representation vector, and realize the cluster mapping symmetry of the communication constraint condition-communication requirement significant interaction fusion representation vector through irreducible low-rank order coefficient manipulation, so as to improve the clustering-based feature representation consistency of the communication constraint condition-communication requirement significant interaction fusion representation vector, thereby improving the accuracy of the optimization instruction obtained by the communication constraint condition-communication requirement significant interaction fusion representation vector through the decoder-based communication parameter optimization controller.

[0045] In particular, the touch screen interface 340 is integrated in the key box and is used to provide elevator status information. The touch screen interface is mainly used to provide elevator status information and enhance the human-computer interaction experience. In one example, the elevator status information may include basic information such as the floor where the elevator is currently located, the running direction (up or down), and the door opening and closing status; and provide an indication of the elevator's operating status, such as whether it is fully loaded, in a fault maintenance mode, etc., to help passengers understand the current status of the elevator.

[0046] As described above, the intelligent construction elevator control system 300 according to the embodiment of the present application can be implemented in various wireless terminals, such as a server with an intelligent construction elevator control algorithm. In a possible implementation, the intelligent construction elevator control system 300 according to the embodiment of the present application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the intelligent construction elevator control system 300 can be a software module in the operating system of the wireless terminal, or can be an application developed for the wireless terminal; of course, the intelligent construction elevator control system 300 can also be one of the many hardware modules of the wireless terminal.

[0047] Alternatively, in another example, the intelligent construction elevator control system 300 and the wireless terminal may also be separate devices, and the intelligent construction elevator control system 300 may be connected to the wireless terminal via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0048] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. Intelligent construction hoist control system, characterized in that: include: The external call box is used to receive the user's call command and transmit the call command to the elevator controller through the wireless module; A key box installed in the elevator car, used to receive control commands from the user; The elevator controller is used to process instructions from the hall call box and the key box to control the speed and direction of the elevator; The elevator controller is further used to dynamically adjust the communication parameters of the LoRa module based on the communication condition information of the area, wherein the communication condition information includes a time queue of a signal strength value, a time queue of a bit error rate, a natural language description of communication requirements, and a device status; Wherein, the elevator controller comprises: A data vectorization module, used for vectorizing the time queue of the signal strength value, the time queue of the bit error rate, the natural language description of the communication demand and the device state to obtain a signal strength time series associated implicit feature vector, a bit error rate time series associated implicit feature vector, a communication demand semantic coding feature vector and a device state semantic embedded coding vector; A communication constraint multimodal information fusion module, used for jointly encoding the signal strength time series associated implicit feature vector, the bit error rate time series associated implicit feature vector and the device state semantic embedded coding vector to obtain a communication constraint condition multimodal fusion representation vector; The feature guided interaction module is used to perform feature guided interaction based on implicit association features on the communication constraint multimodal fusion representation vector and the communication requirement semantic encoding feature vector to obtain a communication constraint-communication requirement significant interaction fusion representation vector.

2. The intelligent construction hoist control system according to claim 1 is characterized in that: The elevator controller further comprises: The communication parameter optimization module is used to generate an optimization instruction for the communication parameters based on the communication constraint condition-communication demand significant interaction fusion representation vector, wherein the optimization instruction includes a recommended SF value.

3. The intelligent construction hoist control system according to claim 2 is characterized in that: The data vectorization module comprises: A time series encoding unit, used for performing time series encoding on the time series of the signal strength value and the time series of the bit error rate respectively using an LSTM model to obtain an implicit feature vector associated with the signal strength time series and an implicit feature vector associated with the bit error rate time series; An embedding coding unit, used for embedding the device state using a device state embedding coding matrix to obtain a device state semantic embedding coding vector; The semantic encoding unit is used to perform semantic embedding encoding on the natural language description of the communication demand using the Bert model to obtain the semantic encoding feature vector of the communication demand.

4. The intelligent construction hoist control system according to claim 3 is characterized in that: The communication constraint multimodal information fusion module is used to: The signal strength time series associated implicit feature vector, the bit error rate time series associated implicit feature vector and the device state semantic embedding coding vector are input into a multimodal joint encoder based on a fully connected layer to obtain the communication constraint condition multimodal fusion representation vector.

5. The intelligent construction hoist control system according to claim 4 is characterized in that: The feature-guided interaction module includes: A condition information construction unit, used to extract the implicit correlation features between the communication constraint condition multimodal fusion representation vector and the communication requirement semantic encoding feature vector as a condition feature vector; A feature modulation unit, configured to perform feature modulation on the communication constraint condition multimodal fusion representation vector and the communication requirement semantic coding feature vector based on the semantic relevance of the communication constraint condition multimodal fusion representation vector and the communication requirement semantic coding feature vector relative to the condition feature vector to obtain a modulated communication constraint condition multimodal fusion representation vector and a modulated communication requirement semantic coding feature vector; A cross-domain interaction coding unit is used to perform cross-domain interaction coding based on a converter structure on the multimodal fusion representation vector of the modulated communication constraint condition, the semantic coding feature vector of the modulated communication requirement, and the conditional feature vector to obtain the communication constraint condition-communication requirement significant interaction fusion representation vector.

6. The intelligent construction hoist control system according to claim 5, characterized in that: The condition information construction unit includes: An implicit correlation feature extraction subunit, used for inputting the communication constraint condition multimodal fusion representation vector and the communication requirement semantic encoding feature vector into an implicit correlation feature capture network to obtain a communication constraint condition-communication requirement implicit correlation feature vector; The feature activation subunit is used to perform feature activation based on the Sigmoid function on the communication constraint condition-communication requirement implicit association feature vector to obtain the conditional feature vector.

7. The intelligent construction hoist control system according to claim 6, characterized in that: The characteristic modulation unit comprises: A first semantic relevance measurement subunit, used to calculate a first semantic relevance factor of the communication constraint condition multimodal fusion representation vector relative to the condition feature vector; A second semantic relevance measurement subunit, used to calculate a second semantic relevance factor of the communication requirement semantic encoding feature vector with respect to the condition feature vector; A weighted modulation subunit is used to normalize the first semantic relevance factor and the second semantic relevance factor, and use the normalized first semantic relevance factor and the second semantic relevance factor to weighted modulate the communication constraint multimodal fusion representation vector and the communication requirement semantic coding feature vector to obtain the modulated communication constraint multimodal fusion representation vector and the modulated communication requirement semantic coding feature vector.

8. The intelligent construction hoist control system according to claim 7, characterized in that: The cross-domain interaction encoding unit is used to: Taking the multimodal fusion representation vector of the post-modulation communication constraint as the query vector, the semantically encoded feature vector of the post-modulation communication requirement as the key vector and the conditional feature vector as the value vector, the multimodal fusion representation vector of the post-modulation communication constraint, the semantically encoded feature vector of the post-modulation communication requirement and the conditional feature vector are input into a feature-to-feature significant guided interaction module based on a converter structure to obtain the communication constraint-communication requirement significant interaction fusion representation vector.

9. The intelligent construction hoist control system according to claim 8, characterized in that: The communication parameter optimization module is used to: The communication constraint condition-communication demand significant interaction fusion representation vector is input into a decoder-based communication parameter optimization controller to obtain the optimization instruction.

10. The intelligent construction hoist control system according to claim 9, characterized in that: A touch screen interface is also included, which is integrated into the key box and is used to provide elevator status information.

Images