An Internet of Things-based collaborative management system and method for hotel self-service devices

Abstract

The present invention relates to the technical field of device management, and discloses an Internet of Things-based collaborative management system and method for hotel self-service devices. The Internet of Things-based collaborative management method for hotel self-service devices includes: collecting communication characteristic data and energy consumption characteristic data of each device, constructing a device communication-energy consumption characteristic mapping model, and calculating a device entropy conversion efficiency index; constructing a network-energy coupling state diagram, and determining a key resource list through threshold screening; generating a multi-dimensional resource allocation plan set through a distributed entropy balance algorithm and a parameter iterative update method. The present invention solves the optimization contradiction between network communication and energy consumption through the above method, realizes the optimization of resource allocation, enables the system to have self-organization and self-adaptation characteristics, significantly improves the energy efficiency ratio, communication quality, resource utilization rate, response speed and fault recovery ability, and provides an efficient and stable network environment for hotel self-service.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment management, and more specifically, to an Internet of Things-based collaborative management system and method for hotel self-service equipment. Background Art

[0002] In the field of hotel equipment management driven by Internet of Things technology, the prior art faces the problem of collaborative optimization of network communication and energy consumption. With the increase in the types and quantities of hotel self-service equipment (including check-in terminals, guest room control systems, food delivery robots, smart door locks, self-service laundry equipment, etc.), the following technical defects have emerged in traditional management systems:

[0003] The communication-energy consumption contradiction is irreconcilable: There is a generally negative correlation between network communication quality and energy consumption in existing systems. High-throughput communication inevitably leads to a sharp increase in equipment energy consumption, while the energy-saving mode will significantly reduce data transmission reliability, forming a "performance-energy consumption" deadlock.

[0004] The energy distribution imbalance is serious: The topological structure and service characteristics of the equipment network lead to non-uniform energy consumption distribution. Key nodes are in an overloaded state for a long time (the energy consumption deviation can reach 40%-60%), while the resource utilization rate of edge nodes is less than 20%, resulting in low overall energy efficiency.

[0005] The lack of dynamic adaptation ability: Traditional threshold control methods perform poorly in dealing with sudden load fluctuations. During network congestion, while the communication quality drops by more than 50%, the energy waste rate increases by 30%-45% instead, forming a vicious cycle of "high consumption and low efficiency".

[0006] The centralized architecture is rigid: The control mode based on the central server has a response delay (usually exceeding 500 ms), and real-time resource allocation cannot be achieved in a complex network environment, resulting in the system adjustment lagging behind environmental changes.

[0007] There are theoretical framework defects in existing technical solutions: The communication optimization problems (such as routing selection, QoS guarantee) and energy management problems (such as power control, load balancing) are treated separately, and independent objective functions are used for optimization. This separated processing method causes the system to fall into a local optimal trap and cannot achieve the global optimal allocation of resources. Therefore, it is urgent to establish a new theoretical framework and technical system to fundamentally solve the problem of collaborative optimization of network communication and energy consumption. Summary of the Invention

[0008] The present invention provides an Internet of Things-based collaborative management system and method for hotel self-service equipment to solve the technical problems in the above related technologies.

[0009] The present invention provides an Internet of Things-based collaborative management method for hotel self-service equipment, including the following steps:

[0010] Collect the communication characteristic data and energy consumption characteristic data of each device in the hotel self-service equipment network, construct a device communication-energy consumption characteristic mapping model, and calculate the device entropy conversion efficiency index;

[0011] Based on the device communication-energy consumption characteristic mapping model and the entropy conversion efficiency index, construct a network-energy coupling state diagram, and determine the key resource list through threshold screening;

[0012] Establish an optimization objective function for the system global entropy conversion efficiency, and generate a multi-dimensional resource allocation scheme set through a distributed entropy balance algorithm and a parameter iterative update method.

[0013] Furthermore, the device communication-energy consumption characteristic mapping model is implemented using a bidirectional neural network structure, including two parts: an encoder and a decoder. The encoder maps communication features to energy consumption features, and the decoder maps energy consumption features to communication features.

[0014] Furthermore, the entropy conversion efficiency index is calculated by the following formula: ;

[0015] Where: Represents the entropy conversion efficiency index of device i; Represents the change in information entropy of device i; Represents the change in energy entropy of device i;

[0016] The information entropy of the device is calculated by the following formula: ;

[0017] Where: Represents the information entropy of device i Represents the probability that the communication state j of device i appears; Represents the sum over index j, and the range of j includes all communication states of device i; Represents the logarithmic function, and the logarithm to the base 2 is used here;

[0018] The energy entropy of the device is calculated by the following formula: ;

[0019] Where: Represents the energy entropy of device i; Is the Boltzmann constant; Is the number of microscopic states of the system; Is the internal energy; Is the equivalent temperature; Represents the natural logarithm.

[0020] Furthermore, in the network-energy coupling state diagram, the node entropy coupling degree is calculated by the following formula:

[0021] ;

[0022] Wherein: represents the node entropy coupling degree of node i; represents the partial derivative of the information entropy of device i with respect to the energy entropy; represents the link weight coefficient; represents the summation over all links connected to node i where E represents the set of all links in the network;

[0023] The link entropy coupling degree is calculated by the following formula: ;

[0024] Wherein: represents the link link entropy coupling degree of; represents the partial derivative of the link information entropy with respect to the energy entropy; represents the link weight coefficient.

[0025] Furthermore, the critical resource list includes a critical node set and a critical link set:

[0026] ;

[0027] Wherein: represents the critical node set; represents the node entropy coupling degree threshold;

[0028] ;

[0029] Wherein: represents the critical link set; represents the link entropy coupling degree threshold.

[0030] Furthermore, the system global entropy conversion efficiency optimization objective function is:

[0031] ;

[0032] Wherein: represents the set of system configuration parameters; represents the maximum value of the parameter set ;

[0033] ;

[0034] ;

[0035] Wherein: and are node weight coefficients; denotes the sum over all nodes \(i\), and \(V\) represents the set of device nodes.

[0036] Furthermore, the distributed entropy balance algorithm transforms the global optimization problem into a local optimization problem, and the calculation formula is as follows:

[0037] ;

[0038] where: denotes the set of neighbor nodes of node \(i\), that is, all nodes directly connected to node \(i\); denotes the influence factor of the information entropy change of node \(j\) on node \(i\); denotes the influence factor of the energy entropy change of node \(j\) on node \(i\); denotes finding the maximum value for the parameter subset ; denotes the sum over all neighbor nodes \(j\) of node \(i\).

[0039] Furthermore, the parameter iterative update method includes introducing an entropy damping factor to adjust the parameter update process:

[0040] ;

[0041] where: is the step size parameter; is the entropy damping factor; is the gradient vector of the local entropy conversion efficiency with respect to the configuration parameters; denotes the configuration parameter of node \(i\) at time \(t + 1\); denotes the configuration parameter of node \(i\) at time \(t\).

[0042] Furthermore, the multi-dimensional resource allocation scheme set includes a routing allocation scheme, a power allocation scheme, and a load allocation scheme. Among them, the optimal communication path determined by the routing allocation scheme satisfies:

[0043] ;

[0044] where: denotes the optimal communication path; denotes finding the path in all path sets that maximizes the following expression; denotes the set of all paths from node \(i\) to node \(j\); denotes the path brings the change in information entropy; denotes the path brings the change in energy entropy;

[0045] The optimal transmission power determined by the power allocation scheme satisfies:

[0046] ;

[0047] Wherein: represents the optimal transmission power; represents finding the power value that maximizes the following expression within the power range ; represents the transmission power of node i; represents the minimum power value; represents the maximum power value; represents the power resulting in the change in information entropy; represents the power resulting in the change in energy entropy;

[0048] The calculation load ratio determined by the load distribution scheme satisfies: ;

[0049] Wherein: represents the load ratio; and represent the entropy conversion efficiency of the node; and represent the entropy coupling degree of the node; represents summing over all nodes j, and V represents the set of all device nodes in the network.

[0050] An Internet of Things-based collaborative management system for hotel self-service devices, which is used to execute the above-mentioned Internet of Things-based collaborative management method for hotel self-service devices, including:

[0051] A device characteristic modeling module, which is used to collect the communication characteristic data and energy consumption characteristic data of each device in the hotel self-service device network, construct a device communication-energy consumption characteristic mapping model, and calculate the device entropy conversion efficiency index;

[0052] A network analysis module, which is used to construct a network-energy coupling state diagram based on the device communication-energy consumption characteristic mapping model and the entropy conversion efficiency index, and determine the key resource list through threshold screening;

[0053] A resource optimization module, which is used to establish an optimization objective function for the global entropy conversion efficiency of the system, and generate a multi-dimensional resource allocation scheme set including communication paths, power control, and load balancing through a distributed entropy balance algorithm and a parameter iteration update method.

[0054] The beneficial effects of the present invention are as follows:

[0055] The information - energy entropy equilibrium theory proposed by the present invention unifies information entropy and physical entropy under the same mathematical framework, establishes a quantitative mapping relationship between information processing and energy consumption, breaks through the limitation of traditional methods that treat communication and energy as independent issues, and fundamentally solves the contradiction that the optimization goals of the two restrict each other.

[0056] The present invention proposes the concept of "entropy conversion efficiency" in the network system, transforms the contradiction between network performance optimization and energy consumption minimization into a system entropy balance problem, and realizes the optimization of resource allocation by maximizing the information entropy gain corresponding to the entropy production per unit energy.

[0057] The present invention establishes a two - way constraint adaptive regulation framework between network performance and energy consumption, making the two no longer a one - way restrictive relationship, but a synergistic relationship of mutual promotion and common optimization. The system can dynamically adjust the resource allocation strategy according to the business importance and network status.

[0058] The distributed entropy balance mechanism designed by the present invention breaks through the traditional centralized control mode, making the system exhibit "self - organizing" and "adaptive" characteristics similar to organisms, achieving global optimization without a central controller, and significantly improving the system response speed and adaptability.

[0059] The energy - information channel self - organization formation algorithm proposed by the present invention realizes the joint optimization of energy efficiency and communication paths in the network, enabling the system to automatically identify and form efficient energy - information transmission channels, maintaining optimal performance in a dynamic environment, and significantly improving the resource utilization efficiency.

[0060] Based on the principle of entropy dynamics, the present invention realizes the spontaneous response and adaptation of the system to environmental changes, no longer relying on predefined rules or simple threshold - triggering mechanisms, and greatly improving the system's fault recovery ability and service availability.

[0061] Through the collaborative optimization of information - energy, the present invention significantly reduces energy consumption while maintaining communication quality, or improves network performance at the same energy consumption level, bringing obvious economic benefits and improved user experience. Brief Description of the Drawings

[0062] Figure 1 It is a flowchart of a method for collaborative management of hotel self - service devices based on the Internet of Things according to the present invention. Detailed Embodiment

[0063] The subject matter described herein will now be discussed with reference to exemplary embodiments. It should be understood that discussing these embodiments is only to enable those skilled in the art to better understand and thus implement the subject matter described herein, and that changes can be made to the functions and arrangements of the elements discussed without departing from the scope of protection of the content of this specification. Each example may omit, substitute, or add various processes or components as needed. Additionally, the features described in some examples may be combined in other examples.

[0064] Embodiment 1, an Internet of Things-based collaborative management method for hotel self-service devices, as Figure 1 shown, includes the following steps:

[0065] Step 101: Receive network operation monitoring data of hotel self-service devices, and generate a device communication - energy consumption characteristic mapping model and an information - energy entropy conversion efficiency index;

[0066] Step 101 includes the following sub-steps:

[0067] Step 101-1, collect dual-domain data of each node device i in the hotel self-service device network, including: communication characteristic data: data transmission rate , communication delay , packet loss rate , signal strength , etc., where t represents the time variable; energy consumption characteristic data: power consumption , voltage / current fluctuation , energy efficiency ratio , etc.;

[0068] Step 101-2, perform preprocessing operations on the collected raw data to obtain a standardized feature vector , where: represents the communication feature sub-vector; represents the energy consumption feature sub-vector;

[0069] Step 101-3, construct a bidirectional characteristic mapping model , which describes the quantitative relationship between the device communication characteristics and the energy consumption characteristics: ;

[0070] This mapping model is implemented using a bidirectional neural network structure and includes: ; ;

[0071] Where: is the encoder function that maps communication features to energy consumption features; is the decoder function that maps energy consumption features to communication features; and are the corresponding model parameter sets;

[0072] Step 101-4, calculate the entropy attribute index of the device based on the bidirectional characteristic mapping model, including:

[0073] Information entropy , representing the uncertainty of the device communication state: ;

[0074] Where: represents the information entropy of device i; represents the probability that the communication state j of device i occurs; represents the summation over index j, and the range of j includes all possible communication states of device i; represents the logarithmic function, and here the logarithm with base 2 is used; Since represents the probability that the communication state j of device i occurs, so takes values in (0, 1), therefore, is negative, then is positive, that is is positive.

[0075] Energy entropy , representing the degree of chaos of the device energy state: ;

[0076] Where: represents the energy entropy of device i; is the Boltzmann constant; is the number of microscopic states of the system; is the internal energy; is the equivalent temperature; represents the natural logarithm;

[0077] Step 101-5, calculate the entropy conversion efficiency index of the device , which quantifies the information entropy gain that can be brought by the change of unit energy entropy: ;

[0078] Where: represents the entropy conversion efficiency index; represents the change amount of the information entropy of device i; represents the change amount of the energy entropy of device i;

[0079] Step 102: Analyze the network topology structure and energy state data, and generate a network-energy coupling state diagram and a key resource list;

[0080] Receive the device characteristic mapping model set generated in Step 101 and the entropy conversion efficiency index set , analyze the network topology structure and energy state data to generate a network - energy coupling state diagram and a list of key resources.

[0081] Step 102 includes the following sub - steps:

[0082] Step 102 - 1, construct the basic network topology graph , where: represents the set of device nodes; represents the set of network connection relationships; for any two nodes , when and only when there is a communication link between them, the connection relationship holds;

[0083] Step 102 - 2, calculate the dual - domain characteristic parameters of the network links. For each link , determine:

[0084] The communication - state characteristic vector ; the energy - consumption state characteristic vector ;

[0085] where: represents the link bandwidth; represents the communication delay; represents the packet - loss rate; represents the power consumption; represents the energy - efficiency ratio;

[0086] Step 102 - 3, calculate the entropy - coupling degree values of network units through an entropy - coupling analyzer, including:

[0087] Node entropy - coupling degree: ;

[0088] where: represents the node entropy - coupling degree of node i; represents the partial derivative of information entropy with respect to energy entropy; represents the link weight coefficient; represents the sum over all links connected to node i, where E represents the set of all links in the network;

[0089] Link entropy - coupling degree: ;

[0090] where: represents the link entropy - coupling degree of link ; represents the partial derivative of link information entropy with respect to energy entropy; represents the link weight coefficient;

[0091] Step 102 - 4, generate the network - energy coupling state diagram , where: represents the set of node entropy coupling degree values; represents the set of link entropy coupling degree values; represents the network - energy coupling state diagram;

[0092] This diagram visually shows the coupling relationship between the network communication structure and the energy distribution, serving as the basis for resource optimization;

[0093] Step 102 - 5, extract the key resource list from the coupling state diagram through a threshold filter, including:

[0094] Set of key nodes: ;

[0095] where: represents the set of key nodes; represents the node entropy coupling degree threshold;

[0096] Set of key links: ;

[0097] where: represents the set of key links; represents the link entropy coupling degree threshold;

[0098] These thresholds are automatically adjusted by analyzing the network historical state data to ensure that the number of key resources is within a reasonable range;

[0099] Step 103: Apply the entropy conversion efficiency optimization algorithm to generate a multi - dimensional resource allocation scheme set;

[0100] Receive the network - energy coupling state diagram generated in Step 102 and the key resource list , and apply the entropy conversion efficiency optimization algorithm to generate a multi - dimensional resource allocation scheme set including communication paths, power control, and load balancing.

[0101] Step 103 includes the following sub - steps:

[0102] Step 103 - 1, establish the system global optimization objective function, which is based on the principle of maximizing entropy conversion efficiency:

[0103]

[0104] where: represents the set of system configuration parameters, including adjustment quantities such as transmission power, routing table, and load ratio; represents the change in the overall information entropy of the system; represents the change in the overall energy entropy of the system; represents taking the maximum value of the parameter set ;

[0105] The change in system entropy is calculated by weighted summation:

[0106] ;

[0107] ;

[0108] where: and are the node weight coefficients; denotes the summation over all nodes i, and V represents the set of device nodes;

[0109] In step 103-2, the distributed entropy balance algorithm is executed to transform the global optimization problem into a set of local sub-problems. For each node i, its local optimization objective is:

[0110]

[0111] where: denotes the set of neighbor nodes of node i, i.e., all nodes directly connected to node i; represents the influence factor of the change in information entropy of node j on node i; represents the influence factor of the change in energy entropy of node j on node i; denotes the subset of configuration parameters that node i can control; denotes the maximum value of the parameter subset ; denotes the summation over all neighbor nodes j of node i;

[0112] In step 103-3, a parameter iterative update method is used to solve the local optimization problem, and the update rule at each time step is:

[0113]

[0114] where: is the step size parameter that controls the update speed; is the gradient vector of the local entropy conversion efficiency with respect to the configuration parameters; denotes the configuration parameter of node i at time t+1; denotes the configuration parameter of node i at time t;

[0115] In step 103-4, an adaptive stabilizer is introduced to adjust the parameter update process through the entropy damping factor :

[0116]

[0117] where: is the step size parameter; is the entropy damping factor; is the gradient vector of the local entropy conversion efficiency with respect to the configuration parameters; represents the configuration parameters of node i at time t+1; represents the configuration parameters of node i at time t;

[0118] The entropy damping factor is automatically calculated based on the node state: ;

[0119] where: represents the entropy damping factor of node i; is the node entropy coupling degree; is the node state fluctuation index, reflecting the stability of the node within a recent time window;

[0120] Step 103-5, based on the optimization results, generate three types of resource allocation schemes:

[0121] Routing allocation scheme: Determine the set of optimal communication paths , where each path satisfies:

[0122] ;

[0123] where: represents the optimal communication path; represents finding the path that maximizes the following expression among all possible path sets ; represents the set of all possible paths from node i to node j; represents the path brings the change in information entropy; represents the path brings the change in energy entropy;

[0124] Power allocation scheme: Determine the set of optimal transmission power values for each node , where each power value satisfies:

[0125] ;

[0126] where: represents the optimal transmission power; represents finding the power value that maximizes the following expression within the power range ; represents the transmission power of node i; represents the minimum power value; represents the maximum power value; represents the power brings the change in information entropy; represents the power brings the change in energy entropy;

[0127] Load distribution scheme: Determine the set of computational load ratios for each node , where each ratio value satisfies:

[0128]

[0129] Where: represents the load ratio; and represent the entropy conversion efficiency of the node; and represent the entropy coupling degree of the node; represents the summation over all nodes j, and V represents the set of all device nodes in the network;

[0130] Step 104: Apply the entropy gravitational field theory and the adaptive maintenance algorithm to construct an energy-information transmission channel network with self-organization and self-repair capabilities.

[0131] Receive the multi-dimensional resource allocation scheme set generated in Step 103 , and apply the entropy gravitational field theory and the adaptive maintenance algorithm to construct an energy-information transmission channel network with self-organization and self-repair capabilities.

[0132] Step 104 includes the following sub-steps:

[0133] Step 104-1, Define the data structure of the dual-domain transmission channel, and each channel records the complete set of transmission attributes:

[0134] Information domain attributes: including bandwidth capacity , transmission delay , link reliability and other parameters;

[0135] Energy domain attributes: including energy consumption level , energy efficiency ratio , temperature gradient and other parameters;

[0136] Step 104-2, Construct an entropy gravitational field model, which uses the entropy conversion efficiency difference as the source of attraction to provide a basis for priority path selection for data traffic:

[0137] Entropy gravitational calculation formula between nodes:

[0138]

[0139] Where: is the system gravitational constant, controlling the overall gravitational strength; is the network topological distance between nodes; is the node function similarity coefficient, with a value range of [0, 1]; and respectively represent the entropy conversion efficiency indicators of node i and node j;

[0140] Step 104-3: Generate the system entropy potential field distribution, which describes the tendency of data flow in the network space:

[0141]

[0142] Among them: represents the position coordinates in the network topology space; represents the distance function from this position to node i; represents the summation over all nodes i; represents the position in the network topology space where the entropy potential field intensity is located;

[0143] Step 104-4: Use the gradient tracking algorithm to form an optimized transmission channel in the entropy potential field, and each channel path satisfies:

[0144]

[0145] Among them: is the preset gradient threshold, which controls the accuracy of channel formation; represents or;

[0146] Step 104-5: Establish a channel health monitoring and adjustment mechanism, including three core components: a health degree evaluator, a channel reconstruction trigger, and a backup channel manager:

[0147] Health degree evaluator: Calculate the channel health status indicator :

[0148]

[0149] Among them: is the health degree evaluation weight coefficient, which satisfies , and is dynamically adjusted according to business requirements; represents the bandwidth capacity; represents the maximum bandwidth; represents the transmission delay; represents the maximum delay; represents the energy efficiency ratio; represents the maximum energy efficiency ratio;

[0150] Channel reconstruction trigger: When the channel health degree is lower than the threshold, start the reconstruction process: If then execute the channel reconstruction;

[0151] Wherein: is the health threshold, which is automatically adjusted by the system according to historical operation data and is different from the node threshold and the link threshold in step 102-5;

[0152] Backup channel manager: maintains an alternate path for critical channels and switches quickly in case of primary channel failure: if a failure is detected, it activates the alternative.

[0153] Step 104-6, implement the load balancing distribution algorithm between channels, and calculate the optimal traffic distribution ratio according to the channel performance parameters:

[0154]

[0155] Wherein: represents the channel entropy conversion efficiency; represents the channel health; represents the summation of all transmission channels in the system;

[0156] The system distributes data traffic accordingly:

[0157]

[0158] Wherein: represents the channel data traffic; represents the channel traffic distribution ratio; represents the total system data traffic.

[0159] The self-maintaining energy-information transmission channel network has the capabilities of self-organization formation, adaptive adjustment, and fault self-recovery, and can maintain the optimal balance between network communication quality and energy efficiency in a dynamic environment.

[0160] The technical effects achieved by this embodiment:

[0161] Based on the information-energy entropy equilibrium theoretical framework, this embodiment effectively solves the optimization contradiction problem between communication quality and energy consumption, the unbalanced network resource allocation problem, the slow system response problem, and the insufficient fault recovery ability problem in the hotel self-service device network. Through core technologies such as entropy conversion efficiency modeling, network-energy coupling state analysis, multi-dimensional resource dynamic allocation, and self-maintaining energy-information transmission channels, this embodiment significantly improves the energy efficiency ratio, communication quality, resource utilization rate, adaptability, and robustness of the hotel self-service device network, providing an efficient, stable, and energy-saving network operation environment for the hotel self-service system, and is particularly suitable for hotel intelligent scenarios with complex communication requirements and limited energy resources.

[0162] An application example of Embodiment 1 is as follows:

[0163] This system was actually deployed in a certain hotel (with 178 guest rooms). The hotel fully deployed a network - energy symbiotic system driven by entropy dynamics and connected a total of 57 Internet of Things self - service device nodes as follows.

[0164] The main challenges faced by this hotel include:

[0165] Network congestion and device response delay problems caused by a sharp increase in the number of customers during the peak business season (October - January of the following year);

[0166] Uneven network coverage in different floors and areas, and weak signals in some dead - end areas resulting in a decline in service quality;

[0167] Unbalanced device energy consumption, where some devices (such as food delivery robots and self - service laundry equipment) still maintain high energy consumption during low - usage periods;

[0168] The traditional centralized network management scheme has a lag in response during peak periods and cannot adjust resource allocation in real - time;

[0169] Device fault handling relies on manual intervention, and the average repair time is too long (about 23.7 minutes);

[0170] The hotel management hopes to solve the above problems through the implementation of this system, especially to reduce energy consumption while improving network service quality, reduce manual intervention, and improve customer satisfaction.

[0171] Example of the implementation process:

[0172] Step 101: Receive the network operation monitoring data of hotel self - service devices and generate an implementation case of the device communication - energy consumption characteristic mapping model and the information - energy entropy conversion efficiency index;

[0173] During the deployment process in this hotel, the system first collected 15 - day baseline data of device operation, sampled every 5 minutes, and formed a device communication - energy consumption characteristic mapping model. The actual data samples of some typical devices are shown in Table 1:

[0174] Table 1: Communication - energy consumption characteristics and entropy conversion efficiency of typical devices

[0175]

[0176] The system uses a customized bidirectional neural network to construct the mapping model. This network has a 5 - layer structure (input layer - 3 hidden layers - output layer), where:

[0177] Input layer: Receive 6 communication characteristic parameters or 4 energy consumption characteristic parameters - Hidden layer: Use a fully - connected neural network with an attention mechanism, and the number of nodes is 12 - 8 - 12 respectively;

[0178] Output layer: Output corresponding energy consumption prediction or communication characteristic prediction;

[0179] Through training with a large amount of data for different time periods (peak period / low period) and different device types, the system has generated a device characteristic mapping model, achieving a prediction error rate of ±6.4% on the validation set, meeting the requirements of practical applications.

[0180] It is particularly worth noting that the system shows significant differences in the entropy conversion efficiency for different types of devices. Smart door locks and smart front desk terminals exhibit relatively high entropy conversion efficiency, while food delivery robots are relatively low. This difference becomes an important basis for subsequent optimal resource allocation.

[0181] Step 102: Analyze the network topology structure and energy state data to generate an implementation case of a network-energy coupling state diagram and a list of key resources;

[0182] Based on the above device characteristic mapping model, the system has constructed a global network-energy coupling state diagram for the hotel. The distribution of the node entropy coupling degree is shown in Table 2:

[0183] Table 2: Distribution of node entropy coupling degree

[0184]

[0185] Based on the set entropy coupling degree thresholds (node threshold , link threshold ), the list of key resources automatically generated by the system is shown in Table 3:

[0186] Table 3: List of key resources

[0187]

[0188] Through in-depth analysis of the network topology structure and energy state, the system has identified several key problem areas in the hotel network: The network link between the lobby and the kitchen has become a communication bottleneck, with an entropy coupling degree as high as 2.8, which is the area that most needs to be optimized first;

[0189] The group of smart door lock devices in the guest room area on the 8th - 12th floors forms an "island" with extremely low energy utilization efficiency in the weak signal area, with an entropy coupling degree less than 0.5;

[0190] Although the number of self-service laundry devices is small, due to their high-power consumption characteristics, they cause significant interference to the surrounding communication network, with a large fluctuation range of entropy coupling degree (0.4 - 2.3);

[0191] Step 103: Apply the entropy conversion efficiency optimization algorithm to generate an implementation case of a multi-dimensional resource allocation plan set;

[0192] Based on the network - energy coupling state diagram and the key resource list, the system executed the entropy conversion efficiency optimization algorithm and generated a multi - dimensional resource allocation plan for the hotel equipment network. The following is the actual situation of the algorithm execution:

[0193] Initial settings: Global optimization goal: Maximize the system entropy conversion efficiency; Node weight: = 0.3 + 0.7·( - ) / ( - ); Step - size parameter: γ = 0.15; Initial value of entropy damping factor: ξ = 0.8; Maximum number of iterations: 200 times;

[0194] Based on the network - energy coupling state diagram and the key resource list, the system executed the entropy conversion efficiency optimization algorithm and generated a multi - dimensional resource allocation plan for the hotel equipment network. The iterative situation of the algorithm execution is shown in Table 4:

[0195] Table 4: Iterative situation of the entropy conversion efficiency optimization algorithm

[0196]

[0197] After 186 rounds of iteration, the algorithm converged to a local optimal solution (the system entropy conversion efficiency increased from 1.28 to 1.98), and the generated multi - dimensional resource allocation plan includes:

[0198] Routing allocation plan: The network communication path table calculated according to the principle of maximizing the entropy conversion efficiency covers 382 main communication links in the hotel. The optimization effects of some key links are shown in Table 5:

[0199] Table 5: Optimization effects of the routing allocation plan

[0200]

[0201] Power allocation plan: The optimal configuration value of the transmission power of network nodes is calculated according to the entropy conversion efficiency. This plan pays special attention to key nodes and realizes dynamic power adjustment. The details are shown in Table 6:

[0202] Table 6: Power allocation plan

[0203]

[0204] Load allocation plan: Determine the computing load ratio of each node, optimize the data processing and computing task allocation in the network, and the optimization results are shown in Table 7:

[0205] Table 7: Optimization results of the load allocation plan

[0206]

[0207] The system converges through iteration, causing the gain in entropy conversion efficiency to gradually slow down until it reaches a stable value, indicating that the resource allocation is approaching the global optimum.

[0208] Step 104: Apply the entropy gravitational field theory and the adaptive maintenance algorithm to construct a case study of an energy-information transmission channel network with self-organization and self-repair capabilities;

[0209] After generating the resource allocation plan, the system constructs an energy-information transmission channel network with self-organization and self-repair capabilities. The following are the actual implementation data for this step:

[0210] Entropy gravitational field model: The system constructs a model using the difference in entropy conversion efficiency as the source of gravity, with the gravitational constant G set to 1.5×10^-3. The actually constructed field strength distribution is shown in Table 8:

[0211] Table 8: Field strength distribution of the entropy gravitational field model

[0212]

[0213] Channel health monitoring and adjustment: The system calculates the channel health status based on the communication quality and energy efficiency ratio, sets the health threshold θH = 0.65, and monitors and maintains the channel health status in real time. The details are shown in Table 9:

[0214] Table 9: Channel health monitoring and adjustment

[0215]

[0216] Load balancing distribution: The system dynamically distributes data traffic based on the channel health and entropy conversion efficiency. The improvement effects compared with the traditional load balancing algorithm are shown in Table 10:

[0217] Table 10: Improvement effects of load balancing distribution

[0218]

[0219] Verification of technical effects:

[0220] Synergistic optimization effect of communication and energy consumption:

[0221] The first core technical effect of this system is to achieve the synergistic optimization of network communication quality and energy consumption, breaking the trade-off relationship between the two in traditional technologies. The comparison data before and after implementation are shown in Table 11:

[0222] Table 11: Comparison of synergistic optimization effects of communication and energy consumption

[0223]

[0224] To verify the stability of this technical effect under different load conditions, the technical team specifically focused on the system performance during peak and off-peak periods. The communication - energy consumption coordination index for different periods is shown in Table 12:

[0225] Table 12: Communication - Energy Consumption Coordination Index for Different Periods

[0226]

[0227] Among them, the communication - energy consumption coordination index is defined as the ratio of the comprehensive score of system communication quality to unit energy consumption, normalized to the interval [0, 1]. The closer it is to 1, the better the coordination effect.

[0228] Performance of the system during a peak holiday season with an occupancy rate > 95%: Even when the system load is 37% higher than the benchmark period, the overall network energy consumption still remains at 72% of the benchmark period level, while the communication quality indicators all exceed the benchmark period by more than 25%.

[0229] System Adaptive and Self - Recovery Capabilities:

[0230] The second core technical effect of this system is to demonstrate excellent adaptive and self - recovery capabilities, significantly enhancing the robustness and availability of the system. The technical team verified this effect in two ways:

[0231] Measurement of Adaptive Capability in Daily Use:

[0232] The adaptive capabilities demonstrated by the system in daily use are shown in Table 13:

[0233] Table 13: Measurement of System Adaptive Capability

[0234]

[0235] Fault Handling and Recovery Capability Testing:

[0236] The technical team conducted 20 simulated fault tests on the system, including scenarios such as node hardware failures, link interruptions, and power fluctuations. The test results are shown in Table 14:

[0237] Table 14: Test Results of Fault Handling and Recovery Capability

[0238]

[0239] The application effect shows that while improving the service quality, this system significantly reduces energy consumption, providing an efficient, stable, and sustainable technical solution for the collaborative management of hotel self-service devices. The entropy dynamics characteristics of the system enable it to not only adapt to the operating pressure during peak periods but also demonstrate excellent resilience and self-healing capabilities in case of failures, which is of great value for enhancing the hotel's intelligence level and customer satisfaction.

[0240] Embodiment 2:

[0241] This embodiment is applied to the collaborative management scenario of hotel self-service devices based on the Internet of Things. On the basis of Embodiment 1, the following technical problems are mainly solved:

[0242] Solving the decision-making delay problem in large-scale device networks. The centralized decision-making architecture in Embodiment 1 forms a computing bottleneck and communication delay when the number of devices increases;

[0243] Solving the problem that different types of devices have different requirements for decision-making timeliness. Embodiment 1 uses a unified decision-making cycle and cannot meet the differentiated needs of devices such as check-in terminals and food delivery robots;

[0244] Solving the problem of underutilized computing resources of edge devices. In Embodiment 1, the calculation of entropy conversion efficiency is mainly completed at the central node;

[0245] Solving the problem of insufficient system robustness in case of network partitioning or communication interruption.

[0246] This embodiment introduces an edge-enhanced hierarchical entropy equilibrium framework. Through an asynchronous decision-making hierarchical collaborative scheduling mechanism, each device node in the network can autonomously optimize at different decision-making levels and time scales while maintaining the goal consistency of global entropy balance.

[0247] Embodiment 2 is optimized and extended on the basis of Embodiment 1. A hotel unmanned check-in method based on face recognition in Embodiment 2 includes the following steps:

[0248] Step 201: Collect device characteristic data and perform timeliness grading;

[0249] Receive the hotel self-service device feature set , where each device feature includes a device identifier, communication frequency data , service priority index , user interaction frequency , energy status time series and computing resource parameters . Classify the devices through a timeliness grading analysis algorithm to generate a device timeliness grading table and a hierarchical decision-making cycle configuration.

[0250] Step 201 includes the following sub-steps:

[0251] Step 201-1, calculate the service response sensitivity of each device :

[0252]

[0253] where: represents the service response sensitivity of device i; is the weight coefficient and satisfies ; represents the communication frequency data of device i; represents the service priority index of device i; represents the user interaction frequency of device i;

[0254] Step 201-2, calculate the energy change rate of each device :

[0255]

[0256] where: represents the energy change rate of device i; represents the length of the sampling time period; represents the sum of all time points t from 1 to ; represents the energy state of device i at time t; represents the energy state of device i at time t-1;

[0257] Step 201-3, calculate the computing power index of each device :

[0258]

[0259] where: represents the computing power index of device i; is the weight coefficient and satisfies ; represents the CPU resource parameter of device i; represents the memory resource parameter of device i; represents the storage resource parameter of device i;

[0260] Step 201-4, comprehensively consider the above three indicators and calculate the comprehensive timeliness index :

[0261]

[0262] where: represents the comprehensive timeliness index of device i; is a weight coefficient and satisfies ; represents the service response sensitivity of device i; represents the energy change rate of device i; represents the computing power index of device i;

[0263] Step 201-5, use the K-means clustering algorithm to classify the devices according to value into three timeliness levels, and assign a corresponding decision period to each level:

[0264] High real-time (Level 1): Decision period = 10 - 100 milliseconds;

[0265] Medium real-time (Level 2): Decision period = 1 - 5 seconds;

[0266] Low real-time (Level 3): Decision period = 30 - 120 seconds;

[0267] Step 202: Use the hierarchical entropy calculation topology generation algorithm to organize the device network into a hierarchical entropy calculation topology structure;

[0268] Receive the device timeliness classification table generated in Step 201 , hierarchical decision period configuration , and the device physical connection diagram

[0269] Step 202 includes the following sub-steps:

[0270] Step 202-1, perform entropy calculation responsibility division, and according to the timeliness level and computing power of the device, assign the entropy calculation responsibility range to each node : ;

[0271] Among them: represents the entropy calculation responsibility range of node i; represents any node j in the device node set V; represents the physical distance between node i and node j; represents the responsibility radius of node i; represents the timeliness level of node j; represents the timeliness level of node i;

[0272] The node responsibility radius is calculated by the following formula: ;

[0273] Wherein: represents the responsibility radius of node i; is a scale parameter adjustable by the system; represents the computing power index of node i; represents the timeliness level of node i;

[0274] Step 202-2: Generate computational dependency edges, calculate the responsibility relationship based on the entropy between nodes, and construct a set of logical computational dependency edges :

[0275] For each pair of nodes , if node is within the entropy calculation responsibility range of node , then add a directed edge from to to to ;

[0276] Step 202-3: Determine the hierarchical boundary nodes and identify the set of boundary nodes connecting different timeliness level layers:

[0277]

[0278] Wherein: represents the set of nodes connecting timeliness level i and level j; represents any node v in the set of device nodes V; represents the timeliness level of node v; represents the entropy calculation responsibility range of node v; represents the timeliness level of node u;

[0279] Step 202-4: Perform topological redundancy optimization. To improve the robustness of the system, add redundant entropy calculation paths for key nodes:

[0280] For each node , calculate the number of its upstream nodes. If this number is less than the preset redundancy threshold , then find the nearest node with sufficient computing power , and add to the responsibility range of ;

[0281] Step 203: Adopt a multi-time-scale asynchronous entropy equilibrium algorithm to decompose the global synchronous optimization into asynchronous optimization processes executed in parallel on multiple time scales, and achieve distributed asynchronous entropy equilibrium optimization;

[0282] Receive the hierarchical entropy calculation topology structure generated in Step 202 and the hierarchical decision cycle configuration and real-time device status data stream ,A multi-time-scale asynchronous entropy balancing algorithm is adopted to decompose the global synchronous optimization into asynchronous optimization processes that are executed in parallel on multiple time scales.

[0283] Step 203 includes the following sub-steps:

[0284] Step 203-1, constructing a hierarchical local entropy model, and constructing a local entropy model suitable for each node according to the timeliness level of the device node:

[0285] Information entropy calculation formula: ;

[0286] in: Represents the information entropy of device node i with timeliness level k; represents the influence weight of node j on node i in level k; represents the communication probability between node i and node j; represents the logarithmic function; represents the sum of all nodes j within the responsibility range of node i;

[0287] Energy entropy calculation formula: ;

[0288] in: represents the energy entropy of device node i with timeliness level k; represents the influence weight of node j on node i in level k; represents the energy consumption required for node i to communicate with node j; represents the decision cycle of the k-th level device; represents the communication energy efficiency ratio; represents the sum of all nodes j within the responsibility range of node i;

[0289] Step 203-2, perform asynchronous optimization scheduling, and perform three levels of optimization calculations according to the timeliness level of the device:

[0290] High real-time device (level 1) optimization objective function:

[0291]

[0292] in: Indicates the routing configuration and power configuration These two parameters seek the values ​​that minimize the objective function; represents the rate of change of information entropy of device i in the high real-time layer over time; represents the rate of change of energy entropy of device i in the high real-time layer over time; Denote the entropy conversion efficiency coefficient of device i in the high real-time layer; Denote the routing configuration; Denote the power configuration;

[0293] Optimization objective function for medium real-time devices (level 2):

[0294]

[0295] Where: Denote the routing configuration and the power configuration Find the values of these two parameters that minimize the objective function; Denote the rate of change of the information entropy of device i in the medium real-time layer over time; Denote the rate of change of the energy entropy of device i in the medium real-time layer over time; Denote the entropy conversion efficiency coefficient of device i in the medium real-time layer; Denote the set of all high real-time devices within the responsibility scope of node i; Denote the influence coefficient; Denote the rate of change of the information entropy of device j in the high real-time layer over time; Denote the summation over all high real-time devices j within the responsibility scope of node i;

[0296] Optimization objective function for low real-time devices (level 3):

[0297]

[0298] Where: Denote the routing configuration , the power configuration and the load distribution Find the values of these three parameters that minimize the objective function; Denote the global influence coefficient of node j; Denote the rate of change of the information entropy of device j in the low real-time layer over time; Denote the rate of change of the energy entropy of device j in the low real-time layer over time; Denote the entropy conversion efficiency coefficient of device j in the low real-time layer; Denote the summation over all device nodes j;

[0299] Step 203-3, Apply state prediction and smoothing update. To coordinate the consistency of decisions at different time scales, introduce state prediction calculation:

[0300]

[0301] Where: Represents the predicted state value of device i at the future time t + Δt; Represents the state value of device i at the current time t; Represents device i at The historical state value at the moment; Represents the prediction weight; Represents the state sampling interval; Represents the number of historical state backtracking steps; Represents the prediction time interval; Represents the sum of all k values from 1 to m;

[0302] Decision smoothing update formula: ;

[0303] Where: Represents the decision value after smoothing update; Represents the timeliness level as The decision value of the device at the current moment; Represents the decision value of the upper timeliness level; Represents the device timeliness level The related smoothing coefficient; Represents the weight coefficient of local decision-making;

[0304] Step 203-4, perform asynchronous message passing, establish three types of message passing channels, and support state synchronization between decision-making units at different time scales:

[0305] Upward message: Transmitted from the low timeliness level to the high timeliness level, including the current state summary and prediction trend; Downward message: Transmitted from the high timeliness level to the low timeliness level, including the optimization strategy and constraint conditions; Emergency message: When it is detected that the state changes rapidly beyond the preset threshold, trigger cross-level emergency synchronization;

[0306] Step 204: Adopt the edge autonomous entropy equilibrium algorithm to enable edge devices to maintain autonomous decision-making ability in case of network anomalies and establish an edge autonomous entropy equilibrium mechanism;

[0307] Receive the hierarchical asynchronous optimization resource allocation scheme set generated in step 203 , the local state of the device , the global state snapshot And the communication status indicator , and adopt the edge autonomous entropy equilibrium algorithm to enable edge devices to maintain autonomous decision-making ability in case of network anomalies.

[0308] Step 204 includes the following sub-steps:

[0309] Step 204-1, perform communication status detection, each device Monitor the connection status with its computationally dependent nodes through a heartbeat detection method:

[0310] Generate a communication status indicator , and the communication status detection period is dynamically adjusted according to the timeliness level of the device. Devices with high real-time performance have a higher detection frequency;

[0311] Step 204-2, create a global state snapshot. During normal communication ( ), each device regularly stores the latest global state snapshot: ;

[0312] Among them: represents the global state snapshot at time t; represents the state value of device j at time t; represents any node j in the device node set V;

[0313] Construct a simplified entropy model: ;

[0314] Among them: represents the simplified entropy model of device i; represents the entropy conversion efficiency of node j; represents the communication energy efficiency ratio between node j and node k; represents the communication probability; represents that both node j and node k are within the responsibility scope of device i;

[0315] Step 204-3, generate an edge autonomous decision. When a communication interruption is detected ( ), calculate the autonomous adjustment boundary: ;

[0316] Among them: represents the autonomous adjustment boundary; represents the initial boundary size; represents the time decay coefficient; represents the current time; represents the time of the most recent global state snapshot; represents the exponential decay function;

[0317] Execute local entropy optimization calculation:

[0318]

[0319] Among them: represents the edge autonomous entropy equilibrium decision set; represents the decision combination that minimizes the objective function for the routing configuration and the power configuration ; Represents the rate of change of the local information entropy of device i over time; Represents the rate of change of the local energy entropy of device i over time; Represents the entropy conversion efficiency coefficient of device i;

[0320] The optimization calculation is subject to the constraint conditions: ;

[0321] Where: Represents the routing configuration of the current calculation; Represents the routing configuration obtained from the last global optimization; Represents the autonomous adjustment boundary;

[0322] Adjust the autonomous boundary parameters:

[0323]

[0324] Where: Represents the autonomous adjustment boundary for the next time period; Represents the autonomous adjustment boundary for the current time period; Represents the boundary adjustment step size; Represents the key performance indicators at the current moment, including communication quality and energy efficiency, etc.;

[0325] Step 204-4, perform state consistency restoration. When communication is restored ( changes from 0 to 1), calculate the decision deviation: ;

[0326] Where: Represents the decision deviation; Represents the edge autonomous decision set; Represents the latest global decision corresponding to the device timeliness level; Represents the norm distance between two decision sets;

[0327] According to the comparison between the deviation magnitude and the preset threshold and , adopt progressive synchronization, fast synchronization or emergency synchronization strategies respectively to ensure a smooth transition of the system state to the globally consistent state;

[0328] The technical achievement of this step is the edge autonomous entropy equilibrium decision set and the autonomous adjustment boundary conditions, which are used to guide the device to make local optimization decisions within a limited range in case of communication interruption, ensuring that the system can still operate normally during network partitioning.

[0329] Step 205: Adopt a multi-level adaptive transmission channel construction algorithm to organize the transmission channels into a hierarchical structure, forming a multi-level adaptive entropy-information transmission channel network.

[0330] Receive the set of hierarchical asynchronous optimization resource allocation schemes generated in step 203 , the set of edge autonomous entropy equilibrium decisions generated in step 204 and the hierarchical entropy calculation topology structure constructed in step 202 , and adopt a multi-level adaptive transmission channel construction algorithm to organize the transmission channels into a hierarchical structure to meet different timeliness requirements.

[0331] Step 205 includes the following sub-steps:

[0332] Step 205-1, define the multi-level channel data structure and create a dedicated transmission channel data structure for each timeliness level: ;

[0333] Among them: represents the set of transmission channels for the timeliness level; represents the set of device nodes any two nodes in; and respectively represent the timeliness levels of node and node ; represents the timeliness level index (1 - high real-time layer, 2 - medium real-time layer, 3 - low real-time layer); represents the transmission channel descriptor;

[0334] The transmission channel descriptor includes the following three types of attributes:

[0335] Information domain attributes: , respectively representing bandwidth capacity, transmission delay, and link reliability;

[0336] Energy domain attributes: , respectively representing energy consumption level, energy efficiency ratio, and temperature gradient;

[0337] Timeliness attributes: , respectively representing refresh frequency, priority, and availability;

[0338] Step 205-2, construct the intra-layer transmission channels, and construct the internal transmission channels for each timeliness level based on the hierarchical asynchronous optimization resource allocation scheme:

[0339] High real-time layer transmission channel ( ): Prioritize low latency and high reliability, and avoid congestion through the reserved resource strategy;

[0340] Medium real-time layer transmission channel ( ): Balance communication quality and energy consumption, and adopt a dynamic resource allocation strategy;

[0341] Low real-time layer transmission channel ( ): Prioritize energy efficiency and reduce the number of transmissions through batch processing and data aggregation;

[0342] Step 205-3, Establish a cross-layer transmission channel. To support data exchange between different timeliness levels, construct a cross-layer transmission channel: ;

[0343] Among them: Represents the set of cross-layer transmission channels; Represents the set of device nodes Any two nodes in; Represents node And node Have different timeliness levels; Indicates that there is a logical calculation dependency between node i and node j; Represents the cross-layer transmission channel descriptor;

[0344] The cross-layer channels are divided into two categories:

[0345] Upward channel: Transmit data from the low timeliness layer to the high timeliness layer, using data summary and event-triggered transmission methods; Downward channel: Transmit data from the high timeliness layer to the low timeliness layer, using policy broadcast and constraint distribution transmission methods;

[0346] Step 205-4, Calculate the hierarchical entropy potential field, and construct an independent entropy potential field model for each timeliness level:

[0347]

[0348] Among them: Represents the entropy potential field value of timeliness level k at position (x, y); Represents the set of all devices with timeliness level k; Represents the information entropy of device i at level k; Represents the energy entropy of device i at level k; Represents the entropy conversion efficiency of device i at level k; Represents the physical location coordinates of device i; Represents the Euclidean distance; Represents the sum of all devices i with timeliness level k;

[0349] The entropy potential fields of each layer have different update periods and transmission characteristics:

[0350] High-real-time layer entropy potential field: The update period is short (millisecond level), mainly driving local fast path adjustment;

[0351] Medium-real-time layer entropy potential field: Obvious regional characteristics, driving collaborative optimization of devices within the range;

[0352] Low-real-time layer entropy potential field: Globally smooth distribution, driving the execution of long-term optimization decisions;

[0353] Step 205-5, implement the multi-time-scale channel maintenance function, and adopt corresponding channel maintenance methods according to the timeliness requirements of different levels:

[0354] High-real-time layer channel maintenance: Adopt local prediction and fast reconstruction methods, and the maintenance period is ;

[0355] Medium-real-time layer channel maintenance: Adopt collaborative detection and regional recovery methods, and the maintenance period is ;

[0356] Low-real-time layer channel maintenance: Adopt comprehensive inspection and resource rebalancing methods, and the maintenance period is ; Cross-layer channel maintenance: Implement the channel status consistency check function, and the maintenance period is ;

[0357] Technical effects of this embodiment:

[0358] While maintaining the information-energy entropy balance theory basis in Embodiment 1, this framework effectively solves the problems of decision-making delay in large-scale device networks, different types of devices' differential requirements for decision-making timeliness, underutilization of edge device computing resources, and insufficient system robustness in case of network partitioning or communication interruption. Through core technologies such as hierarchical timeliness management, asynchronous optimization scheduling, edge autonomous entropy balance, and multi-level transmission channels, this embodiment significantly improves the response speed, service quality, resource utilization efficiency, and system robustness of the hotel self-service device network, and is particularly suitable for complex hotel environments with high requirements for response speed and system robustness.

[0359] The above describes the embodiments of the present invention, but these embodiments are not limited to the above specific implementation manners. The above specific implementation manners are merely illustrative rather than restrictive. Under the inspiration of this embodiment, those of ordinary skill in the art can also make more equivalent embodiments in various forms, all of which fall within the protection scope of this embodiment.

Claims

1. A hotel self-service equipment collaborative management method based on the Internet of Things, characterized in that: The following steps are involved: Collect the communication characteristic data and energy consumption characteristic data of each device in the hotel self-service equipment network, build a device communication-energy consumption characteristic mapping model, and calculate the device entropy conversion efficiency index; Based on the equipment communication-energy consumption characteristic mapping model and entropy conversion efficiency index, a network-energy coupling state diagram is constructed, and a key resource list is determined through threshold screening; Establish the system's global entropy conversion efficiency optimization objective function, and generate a multi-dimensional resource allocation solution set through a distributed entropy balance algorithm and parameter iterative update method; The optimization objective function of the system global entropy conversion efficiency is: ； in: Represents a set of system configuration parameters; Represents a set of parameters Find the maximum value; Indicates the change in the overall information entropy of the system; It represents the change of the overall energy entropy of the system; ； ； in: and is the node weight coefficient; represents the sum of all nodes i, and V represents the set of device nodes; The distributed entropy balance algorithm transforms the global optimization problem into a local optimization problem. The calculation formula is as follows: ； in: Represents the set of neighbor nodes of node i, that is, all nodes directly connected to node i; Represents the impact factor of the information entropy change of node j on node i; represents the impact factor of the energy entropy change of node j on node i; Represents a subset of parameters Find the maximum value; It means summing all neighbor nodes j of node i; and Represent the information entropy changes of devices i and j respectively; and They represent the energy entropy changes of devices i and j respectively; The parameter iterative update method includes introducing an entropy damping factor to adjust the parameter update process: ； in: is the step size parameter; is the entropy damping factor; is the gradient vector of the local entropy conversion efficiency with respect to the configuration parameters; represents the configuration parameters of node i at time t+1; represents the configuration parameters of node i at time t; The multi-dimensional resource allocation scheme set includes a routing allocation scheme, a power allocation scheme, and a load allocation scheme.

2. A hotel self-service equipment collaborative management method based on the Internet of Things according to claim 1, characterized in that: The device communication-energy consumption characteristic mapping model is implemented by a bidirectional neural network structure, and includes two parts, an encoder and a decoder. The encoder maps communication characteristics to energy consumption characteristics, and the decoder maps energy consumption characteristics to communication characteristics.

3. A hotel self-service equipment collaborative management method based on the Internet of Things according to claim 2, characterized in that: The entropy conversion efficiency index is calculated by the following formula: ； in: It represents the entropy conversion efficiency index of device i; represents the change in information entropy of device i; represents the energy entropy change of device i; The information entropy of a device is calculated using the following formula: ； in: represents the information entropy of device i; represents the probability of communication state j of device i occurring; It means summing the index j, where the range of j includes all communication states of device i; represents the logarithmic function, where the logarithm with base 2 is used; The energy entropy of a device is calculated using the following formula: ； in: represents the energy entropy of device i; is the Boltzmann constant; is the number of microscopic states of the system; is internal energy; is the equivalent temperature; Represents the natural logarithm.

4. A hotel self-service equipment collaborative management method based on the Internet of Things according to claim 3, characterized in that: In the network-energy coupling state diagram, the node entropy coupling degree is calculated by the following formula: ； in: represents the node entropy coupling degree of node i; represents the partial derivative of the information entropy of device i with respect to the energy entropy; represents the link weight coefficient; Represents all links connected to node i Sum, where E represents the set of all links in the network; The link entropy coupling is calculated by the following formula: ； in: Indicates link Link entropy coupling degree; represents the partial derivative of link information entropy with respect to energy entropy; Represents the link weight coefficient.

5. A hotel self-service equipment collaborative management method based on the Internet of Things according to claim 4, characterized in that: The key resource list includes a key node set and a key link set: ； in: Represents a set of key nodes; Indicates the node entropy coupling threshold; Represents a collection of device nodes; ； in: represents a set of critical links; Indicates the link entropy coupling threshold; Represents a collection of network connection relationships.

6. A hotel self-service equipment collaborative management method based on the Internet of Things according to claim 5, characterized in that: The optimal communication path determined by the routing allocation scheme satisfies: ； in: represents the optimal communication path; Indicates that in all path sets Find the path that maximizes the following expression; represents the set of all paths from node i to node j; Indicates the path The amount of change in information entropy brought about; Indicates the path The amount of energy entropy change brought about; The optimal transmission power determined by the power allocation scheme satisfies: ； in: represents the optimal transmission power; Indicates the power range Find the power value that maximizes the following expression; represents the transmission power of node i; Indicates the minimum power value; Indicates the maximum power value; Indicates power The amount of change in information entropy brought about; Indicates power The amount of energy entropy change brought about; The calculated load ratio determined by the load distribution scheme satisfies: ； in: Indicates load ratio; and Indicates the entropy conversion efficiency of the node; and Indicates the entropy coupling degree of the node; represents the sum of all nodes j, and V represents the set of all device nodes in the network.

7. A hotel self-service equipment collaborative management system based on the Internet of Things, characterized in that: It is used to execute a hotel self-service equipment collaborative management method based on the Internet of Things as described in any one of claims 1 to 6, comprising: The equipment characteristic modeling module is used to collect the communication characteristic data and energy consumption characteristic data of each device in the hotel self-service equipment network, build the equipment communication-energy consumption characteristic mapping model, and calculate the equipment entropy conversion efficiency index; A network analysis module, for constructing a network-energy coupling state diagram based on the device communication-energy consumption characteristic mapping model and the entropy conversion efficiency index, and determining a key resource list through threshold screening; The resource optimization module is used to establish the system's global entropy conversion efficiency optimization objective function, and generate a multi-dimensional resource allocation solution set including communication path, power control and load balancing through a distributed entropy balancing algorithm and parameter iterative update method.

Images