Internet of Things generator with remote visual control mechanism and state energy consumption prediction method thereof

Abstract

The invention discloses an Internet of Things generator with a remote visual control mechanism and a state energy consumption prediction method thereof.The Internet of Things generator comprises a generator body, and the bottom end of the generator body is fixedly provided with a first generator shell and a second generator shell. The method comprises the following steps: monitoring and analyzing data, finding a fault problem in advance, timely sending an alarm notice when a fault occurs, reproducing a fault site to provide a basis for maintenance, providing maintenance reminding, prolonging the service life of equipment, automatically generating an equipment report, managing and storing historical data, realizing stable and rapid communication, never losing line, providing bidding management, and sharing industry resources. Normal operation of the Internet of Things generator is ensured, and the power generation efficiency of the Internet of Things generator is improved; and by arranging the remote control mechanism, related information of the Internet of Things generator is transmitted to a background, so that background personnel can know the information of the Internet of Things generator in time and carry out overhaul and maintenance.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet of Things generators, and in particular to an Internet of Things generator with a remote visualization control mechanism and a state energy consumption prediction method thereof. Background Art

[0002] The IoT generator has built-in sensors and monitoring systems that can monitor the current electricity demand in real time. These sensors can measure parameters such as the power demand and voltage demand of power-consuming equipment. According to the monitored power demand, the IoT generator will automatically adjust the power generation. When the power demand increases, the IoT generator will automatically increase the power generation; when the power demand decreases, the power generation will be reduced. By dynamically adjusting the power generation, the IoT generator can meet the real-time power demand. The IoT generator also has the ability to generate power optimally. It can calculate the optimal power generation and power generation method based on the current power demand and the performance characteristics of the generator, so as to avoid excess or insufficient power generation, thereby improving the utilization efficiency of electricity. The IoT generator can choose the appropriate power generation method according to the actual situation. It can choose the best power generation method based on the available energy resources, such as fossil fuels, solar energy, wind energy, etc., and adjust it in real time when the power demand changes.

[0003] However, traditional IoT generators have the following disadvantages: When a traditional IoT generator is generating electricity, the user cannot understand the operating status of the IoT generator. When a IoT generator fails, the user cannot maintain it in time, which reduces the power generation efficiency of the IoT generator. Summary of the invention

[0004] The purpose of the present invention is to provide an Internet of Things generator with a remote visualization control mechanism and a state energy consumption prediction method thereof, so as to solve the problem that during the power generation process of the traditional Internet of Things generator proposed in the above background technology, the user cannot understand the operating status of the Internet of Things generator, and when the Internet of Things generator fails, the user cannot maintain it in time, which reduces the power generation efficiency of the Internet of Things generator.

[0005] To achieve the above-mentioned purpose, the present invention provides the following technical solutions: an Internet of Things generator with a remote visualization control mechanism, comprising a generator body, a first generator casing fixedly installed at the bottom end of the generator body, a second generator casing provided at the top end of the first generator casing, a mounting mechanism fixedly installed on one side of the second generator casing, the mounting mechanism comprising a mounting base and two height platforms, both sides of the top end of the mounting base are respectively fixedly connected to the bottom ends of the two height platforms, a synchronization rod is provided at the top end between the two height platforms, a visualization screen is fixedly installed at the middle part of the synchronization rod, a remote control mechanism is fixedly installed at the top end of the visualization screen, the remote control mechanism comprises a transmitter, a transmitter, a controller, a receiver and a receiver, and the transmitter, the transmitter, the receiver and the receiver are all connected to the controller.

[0006] As a preferred technical solution of the present invention, one side of the two height platforms is rotatably connected with a direction rod, and the ends of the two direction rods away from the height platforms are rotatably connected with the two ends of the synchronization rod respectively, and a turntable is installed at the connection between the two height platforms and the two direction rods, and one side of the two turntables is fixedly installed with a cooling plate, and one side of the two cooling plates is contacted and connected with the side directly opposite to the synchronization rod, and a connecting rod is installed between the two turntables. After the direction rod rotates along the height platform, the direction rod drives the adjacent direction rods to move synchronously through the synchronization rod to adjust the position of the visualization screen. During the rotation of the direction rod, the turntable is driven to move synchronously, and the turntable drives the cooling plate to move synchronously. The cooling plate is in contact with the synchronization rod, and heat is transferred and dissipated from the back of the visualization screen.

[0007] As a preferred technical solution of the present invention, one side of the mounting base is fixedly connected to a side directly opposite to the second generator casing, and the mounting mechanism is fixed to the second generator casing via the mounting base.

[0008] As a preferred technical solution of the present invention, the sending component includes an input module, a first control module, a first storage module, a recording module, a microphone, an A / D converter and a wireless signal generating module. The recording module is connected to the microphone, and the microphone is connected to the A / D converter. The A / D converter, the wireless signal generating module, the input module and the first storage module are all connected to the first control module. The microphone receives the instructions recorded by the user through the recording module. The received instructions are encrypted, filtered and converted by the A / D converter. After the signal conversion, it is transmitted to the first control module. The first control module transmits the signal to the controller. The wireless signal generating module generates a corresponding sending signal and transmits it to the first control module. The first storage module stores the sending signal.

[0009] As a preferred technical solution of the present invention, the receiving component includes a playback module, a speaker, a D / A converter, a wireless signal receiving module, a second storage module, a second control module and a driving module. The playback module is connected to the speaker, and the speaker is connected to the D / A converter. The D / A converter, the wireless signal receiving module, the second storage module and the driving module are all connected to the second control module. The speaker receives the receiving signal transmitted by the playback module, and the receiving signal is decrypted, filtered and converted by the D / A converter. After the signal conversion, it is transmitted to the second control module, and the second control module transmits the signal to the controller, and the wireless signal receiving module generates a corresponding receiving signal and transmits it to the second control module, and the second storage module stores the received signal.

[0010] As a preferred technical solution of the present invention, the visualization screen includes a sensor data display module, an input status display module, a relay status display module, an engine data display module, a power generation data display module, a load data display module and a fault data display module. The sensor data display module displays the sensor data in the Internet of Things generator, the input status display module displays the input status of the connector on the Internet of Things generator, the relay status display module displays the relay status on the Internet of Things generator, the engine data display module displays the engine data on the Internet of Things generator, the power generation data display module displays the power generation data of the Internet of Things generator, the load data display module displays the load data of the Internet of Things generator, and the fault data display module displays the fault data of the Internet of Things generator. The sensor data display module displays the sensor data in the Internet of Things generator, the input status display module displays the input status of the connector on the Internet of Things generator, the relay status display module displays the relay status on the Internet of Things generator, the engine data display module displays the engine data on the Internet of Things generator, the power generation data display module displays the power generation data of the Internet of Things generator, the load data display module displays the load data of the Internet of Things generator, and the fault data display module displays the fault data of the Internet of Things generator.

[0011] As a preferred technical solution of the present invention, a plurality of first heat dissipation holes are provided on the surface of the first generator casing and the surface of the second generator casing, a heat dissipation cover is mounted on the top end of the second generator casing, and a plurality of second heat dissipation holes are provided on the surface of the heat dissipation cover. The first heat dissipation holes and the second heat dissipation holes are provided to facilitate the contact between the high-temperature gas generated by the generator body during operation and the gas in the external environment for heat transfer, heat dissipation and cooling.

[0012] As a preferred technical solution of the present invention, a fixing plate is fixedly installed at one end of the first generator casing, and the generator body is fixed at the installation position through the fixing plate.

[0013] As a preferred technical solution of the present invention, an assembly plate is fixedly installed on one side of the first generator casing and one side of the second generator casing, and bolts are threadedly connected between the two assembly plates. The user tightens the bolts to assemble the two assembly plates together, and the first generator casing and the second generator casing are assembled together through the assembly plate.

[0014] As a preferred technical solution of the present invention, an interface is provided on one side of the heat dissipation cover.

[0015] A method for predicting energy consumption of an IoT generator state with a remote visualization control mechanism, the specific process is as follows: Step 1: Install sensors to collect real-time status data of the generator and related information transmitted by the remote visualization control mechanism. The sensors used and their installation locations are as follows: Temperature sensor: installed inside the generator body to monitor the operating temperature of the generator; the temperature sensor may be installed on the inner surface of the first generator casing (2) and the second generator casing (3); Voltage sensor and current sensor: installed in the generator output circuit to monitor the voltage and current of the generator; these sensors can be installed at the circuit connection or directly connected to the power line; Vibration sensor: installed on the generator casing surface or bearing to monitor the vibration of the generator; the vibration sensor can help detect whether the generator is operating abnormally; Microphone: used for collecting sound of the recording module (1224) so ​​as to monitor the sound signal during the operation of the generator and to diagnose the operating status of the generator; The required data includes the real-time status data of the generator, including the generator operating status, temperature, voltage, current and other related parameters as well as historical load demand data and energy supply data; Parameter variables Normal Operation Standby Shutdown Fault Internal Temperature of Generator Generator external temperature (External Temperature) Phase A Voltage Phase B Voltage Phase C Voltage Total Voltage Phase A Current Phase B Current Phase C Current Total Current Average Load Demand Peak Load Demand Load Demand Distribution Electricity Generation Energy Consumption Energy Sources Step 2: Establish a deep neural network model combining convolutional neural network (CNN) and long short-term memory network (LSTM) based on sensor data; The CNN structure is used to process the generator status data: Input data dimensions: The dimensions of the generator status data are (N, C, H, W), where: N represents the number of samples; C represents the number of channels, which represents the number of different types of sensors; H represents the height of the feature map, which can be the time step or spatial resolution; W represents the width of the feature map, which represents the number of features or the length of the time series: Convolutional Layer: The first convolution layer: 64 3x3 convolution kernels, stride 1, using ReLU activation function; The second convolution layer: 128 3x3 convolution kernels, stride 1, using ReLU activation function; The third convolution layer: 256 3x3 convolution kernels, stride 1, using ReLU activation function; Pooling layer: Each convolutional layer is followed by a 2x2 max pooling layer to reduce the size of the feature map; The LSTM structure is used to process the time series data of the remote control mechanism: Input Layer: Input data dimensions: The dimensions of the remote control mechanism data are (N, L, D), where: N represents the number of samples; L represents the length of the time series; D represents the dimension of the feature, which can be the number of sensors or the number of control parameters: Multi-layer LSTM layers: First layer LSTM: number of units: 64, activation function: tanh; Second layer LSTM: number of units: 64, activation function: tanh; The third layer LSTM: number of units: 32, activation function: tanh; Fully connected layer: Fully connected layer 1: Number of neurons: 128, activation function: ReLU; Fusion and fully connected layers: Fusion layer: The outputs of CNN and LSTM can be fused by concatenation operation; Fully connected layer: The first fully connected layer has 1024 neurons and uses the ReLU activation function. The second fully connected layer has 512 neurons and uses the ReLU activation function. In a multi-layer LSTM, each layer of LSTM receives the hidden state of the previous layer as input. The specific process is: The forget gate is used to control the forgetting of part of the information in the memory cell of the previous time step: ; in is the output of the forget gate, is the weight matrix of the forget gate, is the hidden state of the previous time step, is the input of the current time step, is the bias term of the forget gate, σ is the Sigmoid activation function; The input gate controls the amount of new information that enters the memory cell: ; in is the input gate output, is the weight matrix of the input gate, is the hidden state of the previous time step, is the input of the current time step, is the bias term of the input gate, σ is the Sigmoid activation function; The new information is used to update the contents of the memory cell: ; in For new information, To calculate the weight matrix of new information, is the hidden state of the previous time step, is the input of the current time step, is the bias term of the new information, tanh is the hyperbolic tangent function, and is the activation function of the hidden layer; Update memory cells and calculate new memory cell states: ; in is the state of memory cells, is the output of the forget gate, is the memory cell state at the previous time step, is the input gate output, For new information; The output gate controls the information in the hidden state to be passed to the next time step: ; in is the output gate output, is the weight matrix of the output gate, is the hidden state of the previous time step, is the input of the current time step, is the bias term of the output gate; Hide status updates: ; is the hidden state at the current time step, is the output gate output, is the memory cell state at the current time step, tanh is the hyperbolic tangent function, and is the activation function of the hidden layer; Step 3: Use the back propagation algorithm to train the model and optimize the model parameters. The specific process of the back propagation algorithm is to calculate the output of the convolutional neural network CNN. And the output of the long short-term memory network LSTM ; Combine the two to get the fused feature table ; Use the fully connected layer for the final prediction output ; Use the mean square error MSE loss function to measure the prediction results and real results The difference between: ; Back propagation calculates the loss function with respect to the predicted output The gradient is: ; According to the chain rule, the gradient of the loss function is passed back to the fusion layer: ; Pass the gradients of the fusion layer back to each network: ; According to the gradient of each network, the back propagation algorithm is used to update the parameters of each network; According to the gradient calculated by the back-propagation algorithm, the parameters of the neural network model are updated using stochastic gradient descent to minimize the loss function. Step 4, divide the data set: divide the data set into training set, validation set and test set; use the status data of the generator and the information of the remote control mechanism as the input of the model, and use the predicted energy consumption as the output of the model; use the back propagation algorithm to train the model and optimize the model parameters so that it can accurately predict the future energy consumption; Depending on the input from different sensors, the following predictions may exist: 1. Temperature sensor data: Forecast under normal temperature conditions: It is predicted that the internal and external temperatures of the generator will remain within the normal range in the next few hours, and the energy consumption will be relatively stable; Prediction of abnormal temperature fluctuations: If the temperature fluctuates abnormally, it may indicate that the generator is faulty or the external environment has changed. The model will predict the trend of future temperature fluctuations and provide early warning information; 2. Voltage sensor and current sensor data: Prediction of voltage and current stability: If the voltage and current data remain stable, the model will predict the future operating status and energy consumption of the generator to maintain stable power generation; Prediction of voltage or current anomalies: If the voltage or current is abnormal, it may indicate that the output power of the generator is unstable or there is a fault. The model will predict the future working status and provide corresponding early warning information; 3. Vibration sensor data: Prediction under normal vibration conditions: predict that the vibration level of the generator will remain within the normal range and the working state will be stable in the future; Prediction of abnormal vibration: If the vibration data shows abnormalities, it may indicate problems with the generator’s bearings or other components. The model will predict future vibration conditions and provide fault diagnosis information; 4. Microphone data: Prediction under normal working conditions: predict that the sound signal during the future operation of the generator will remain within the normal range without abnormal conditions; Prediction of abnormal sound: If the microphone data detects abnormal sound, it may indicate abnormal operation or failure of the generator. The model will predict the trend of future sound signal changes and provide corresponding early warning information; Step 5: Deploy the trained model to the remote visualization control mechanism in the IoT generator system; monitor the status data of the generator and the remote control mechanism information in real time and use them as the input of the model; use the trained model to predict future energy consumption and display the prediction results on the visualization screen to provide timely monitoring and decision support for operators.

[0016] Compared with the prior art, the present invention has the following beneficial effects: 1. By setting up a visual screen, the equipment operation status can be monitored and recorded anytime and anywhere, data can be monitored and analyzed, faults can be discovered in advance, and when a fault occurs, an alarm notification can be issued in time, and the fault site can be reproduced to provide a basis for maintenance, maintenance reminders can be provided, and the equipment life can be extended. Equipment reports can be automatically generated, and historical data can be stored. The communication is stable and fast, and it will never be disconnected. It provides bidding management and shares industry resources to ensure the normal operation of the IoT generator and improve the power generation efficiency of the IoT generator. 2. By setting up a remote control mechanism, the microphone receives the instructions recorded by the user through the recording module. The received instructions are encrypted, filtered and converted through the A / D converter. After the signal conversion, it is transmitted to the first control module. The speaker receives the receiving signal transmitted by the playback module. The receiving signal is decrypted, filtered and converted through the D / A converter, which is convenient for the background personnel to timely understand the information of the IoT generator and perform inspection and maintenance.

[0017] 3. The method for predicting the state energy consumption of an IoT generator with a remote visualization control mechanism of the present invention is based on a deep neural network, and combines a convolutional neural network (CNN) and a long short-term memory network (LSTM) to analyze the real-time state of the generator and predict future energy consumption. It can simultaneously predict the normal and abnormal temperature conditions, the normal and abnormal voltage or current conditions, the abnormal vibration conditions, and the abnormal sound conditions, and display the prediction results on a visualization screen, providing timely monitoring and decision support for operators. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 is a side view of the present invention; Figure 2 It is a partial schematic diagram of the present invention; Figure 3 It is a connection diagram of the mounting mechanism of the present invention and the visualization screen; Figure 4 It is a schematic diagram of the architecture of the remote control mechanism of the present invention; Figure 5 It is a schematic diagram of the structure of the visualization screen of the present invention; Figure 6 It is a rear view of the generator body of the present invention; Figure 7is a top view of the generator body of the present invention; Figure 8 It is a flow chart of the present invention.

[0019] In the figure: 1. Generator body; 2. First generator housing; 3. Second generator housing; 4. Fixing plate; 5. Mounting mechanism; 51. Mounting base; 52. Height platform; 53. Turntable; 54. Connecting rod; 55. Direction rod; 56. Synchronous rod; 57. Refrigeration plate; 6. Heat dissipation cover; 7. Assembly plate; 8. Bolt; 9. First heat dissipation hole; 10. Second heat dissipation hole; 11. Visualization screen; 111. Sensor data display module; 112. Input status display module; 113. Relay status display module; 114. Engine data display module; 115. Power generation data display module; 116. Load data display module; 117. Fault data display module; 12, remote control mechanism; 121, transmitter; 122, sending element; 1221, input module; 1222, first control module; 1223, first storage module; 1224, recording module; 1225, microphone; 1226, A / D converter; 1227, wireless signal generating module; 123, controller; 124, receiver; 125, receiving element; 1251, playing module; 1252, speaker; 1253, D / A converter; 1254, wireless signal receiving module; 1255, second storage module; 1256, second control module; 1257, driving module; 13, interface. DETAILED DESCRIPTION

[0020] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] See also Figure 1-7The present invention provides an Internet of Things generator with a remote visualization control mechanism, comprising a generator body 1, a first generator casing 2 is fixedly installed at the bottom end of the generator body 1, a second generator casing 3 is provided at the top end of the first generator casing 2, a mounting mechanism 5 is fixedly installed on one side of the second generator casing 3, the mounting mechanism 5 comprises a mounting base 51 and two height platforms 52, both sides of the top end of the mounting base 51 are respectively fixedly connected to the bottom ends of the two height platforms 52, a synchronization rod 56 is provided at the top end between the two height platforms 52, a visualization screen 11 is fixedly installed at the middle part of the synchronization rod 56, a remote control mechanism 12 is fixedly installed at the top end of the visualization screen 11, the remote control mechanism 12 comprises a transmitter 121, a transmitter 122, a controller 123, a receiver 124 and a receiver 125, and the transmitter 121, the transmitter 122, the receiver 124 and the receiver 125 are all connected to the controller 123.

[0022] One side of the two height platforms 52 is rotatably connected with a direction rod 55, and one end of the two direction rods 55 away from the height platform 52 is rotatably connected to the two ends of the synchronization rod 56 respectively. A turntable 53 is installed at the connection between the two height platforms 52 and the two direction rods 55, and a cooling plate 57 is fixedly installed on one side of the two turntables 53. One side of the two cooling plates 57 is in contact and connected with the side directly opposite to the synchronization rod 56. A connecting rod 54 is installed between the two turntables 53. After the direction rod 55 rotates along the height platform 52, the direction rod 55 drives the adjacent direction rods 55 to move synchronously through the synchronization rod 56 to adjust the position of the visualization screen 11. During the rotation of the direction rod 55, it drives the turntable 53 to move synchronously, and the turntable 53 drives the cooling plate 57 to move synchronously. The cooling plate 57 contacts the synchronization rod 56 to transfer and dissipate heat from the rear of the visualization screen 11.

[0023] One side of the mounting base 51 is fixedly connected to the side of the second generator casing 3 that is directly opposite to the second generator casing 3 , and the mounting mechanism 5 is fixed to the second generator casing 3 through the mounting base 51 .

[0024] The sending component 122 includes an input module 1221, a first control module 1222, a first storage module 1223, a recording module 1224, a microphone 1225, an A / D converter 1226 and a wireless signal generating module 1227. The recording module 1224 is connected to the microphone 1225, and the microphone 1225 is connected to the A / D converter 1226. The A / D converter 1226, the wireless signal generating module 1227, the input module 1221 and the first storage module 1223 are all connected to the first control module 1222. The microphone 1225 receives the instructions recorded by the user through the recording module 1224. The received instructions are encrypted, filtered and converted by the A / D converter 1226. After the signal conversion, it is transmitted to the first control module 1222. The first control module 1222 transmits the signal to the controller 123. The wireless signal generating module 1227 generates a corresponding sending signal and transmits it to the first control module 1222. The first storage module 1223 stores the sending signal.

[0025] The receiving component 125 includes a playing module 1251, a speaker 1252, a D / A converter 1253, a wireless signal receiving module 1254, a second storage module 1255, a second control module 1256 and a driving module 1257. The playing module 1251 is connected to the speaker 1252, the speaker 1252 is connected to the D / A converter 1253, the D / A converter 1253, the wireless signal receiving module 1254, the second storage module 1255 and the driving module 1257 are all connected to the second control module 1256. The speaker 1252 receives the receiving signal transmitted by the playing module 1251, the receiving signal is decrypted, filtered and converted by the D / A converter 1253, and the signal is transmitted to the second control module 1256 after the signal conversion. The second control module 1256 transmits the signal to the controller 123, and the wireless signal receiving module 1254 generates a corresponding receiving signal and transmits it to the second control module 1256, and the second storage module 1255 stores the receiving signal.

[0026] The visualization screen 11 includes a sensor data display module 111, an input status display module 112, a relay status display module 113, an engine data display module 114, a power generation data display module 115, a load data display module 116 and a fault data display module 117. The sensor data display module 111 displays the sensor data in the Internet of Things generator, the input status display module 112 displays the input status of the connector on the Internet of Things generator, the relay status display module 113 displays the relay status on the Internet of Things generator, the engine data display module 114 displays the engine data on the Internet of Things generator, the power generation data display module 115 displays the power generation data of the Internet of Things generator, the load data display module 116 displays the load data of the Internet of Things generator, and the fault data display module 117 displays the fault data of the Internet of Things generator.

[0027] A plurality of first heat dissipation holes 9 are provided on the surface of the first generator casing 2 and the surface of the second generator casing 3. A heat dissipation cover 6 is mounted on the top of the second generator casing 3. A plurality of second heat dissipation holes 10 are provided on the surface of the heat dissipation cover 6. The first heat dissipation holes 9 and the second heat dissipation holes 10 are provided to facilitate the contact between the high-temperature gas generated by the generator body 1 during operation and the gas in the external environment for heat transfer, heat dissipation and cooling.

[0028] A fixing plate 4 is fixedly mounted on one end of the first generator casing 2 , and the generator body 1 is fixed at the mounting position through the fixing plate 4 .

[0029] An assembly plate 7 is fixedly installed on one side of the first generator casing 2 and one side of the second generator casing 3. A bolt 8 is threadedly connected between the two assembly plates 7. The user tightens the bolt 8 to assemble the two assembly plates 7 together. The first generator casing 2 and the second generator casing 3 are assembled together through the assembly plate 7.

[0030] An interface 13 is formed on one side of the heat dissipation cover 6 .

[0031] In the present invention, the user screws the bolt 8 to assemble the two assembly plates 7 together, and the first generator casing 2 and the second generator casing 3 are assembled together through the assembly plate 7. After the direction rod 55 rotates along the height platform 52, the direction rod 55 drives the adjacent direction rods 55 to move synchronously through the synchronization rod 56 to adjust the position of the visualization screen 11. During the rotation of the direction rod 55, the turntable 53 is driven to move synchronously, and the turntable 53 drives the cooling plate 57 to move synchronously. The cooling plate 57 contacts the synchronization rod 56, and heat is transferred and dissipated from the back of the visualization screen 11. The first heat dissipation hole 9 and the second heat dissipation hole 10 are provided to facilitate the contact between the high-temperature gas generated by the generator body 1 during operation and the gas in the external environment for heat transfer, heat dissipation and cooling. The visualization screen 11 displays the sensor data, input status, relay status, engine data, power generation data, load data and fault data of the generator body 1, and transmits the information to the background through the transmitter 121, and receives the instructions transmitted from the background through the receiver 124. The microphone 1225 receives the user's recording module 1224 records the instruction, the received instruction is encrypted, filtered and converted by the A / D converter 1226, and the signal is transmitted to the first control module 1222 after the signal conversion, the first control module 1222 transmits the signal to the controller 123, the wireless signal generating module 1227 generates a corresponding transmission signal and transmits it to the first control module 1222, the first storage module 1223 stores the transmission signal, the transmitter 121 sends the to-be-sent signal in the controller 123 to the background, the speaker 1252 receives the received signal transmitted by the playing module 1251, the received signal is decrypted, filtered and converted by the D / A converter 1253, and the signal is transmitted to the second control module 1256 after the signal conversion, the second control module 1256 transmits the signal to the controller 123, and the wireless signal receiving module 1254 generates a corresponding reception signal and transmits it to the second control module 1256, the second storage module 1255 stores the reception signal, and the receiver 124 transmits the reception signal in the controller 123 to the sensor in the generator body 1.

[0032] A method for predicting the state energy consumption of an IoT generator with a remote visualization control mechanism is proposed. Based on a deep neural network, a convolutional neural network (CNN) and a long short-term memory network (LSTM) are combined to analyze the real-time state of the generator and predict future energy consumption. The specific process is as follows: Step 1: Install sensors to collect real-time status data of the generator and related information transmitted by the remote visualization control mechanism. The sensors used and their installation locations are as follows: Temperature sensor: installed inside the generator body to monitor the operating temperature of the generator; the temperature sensor may be installed on the inner surface of the first generator casing (2) and the second generator casing (3); Voltage sensor and current sensor: installed in the generator output circuit to monitor the voltage and current of the generator; these sensors can be installed at the circuit connection or directly connected to the power line; Vibration sensor: installed on the generator casing surface or bearing to monitor the vibration of the generator; the vibration sensor can help detect whether the generator is operating abnormally; Microphone: used for collecting sound of the recording module (1224) so ​​as to monitor the sound signal during the operation of the generator and to diagnose the operating status of the generator; The required data includes the real-time status data of the generator, including the generator operating status, temperature, voltage, current and other related parameters as well as historical load demand data and energy supply data; Parameter variables Normal Operation Standby Shutdown Fault Internal Temperature of Generator Generator external temperature (External Temperature) Phase A Voltage Phase B Voltage Phase C Voltage Total Voltage Phase A Current Phase B Current Phase C Current Total Current Average Load Demand Peak Load Demand Load Demand Distribution Electricity Generation Energy Consumption Energy Sources Step 2: Establish a deep neural network model combining convolutional neural network (CNN) and long short-term memory network (LSTM) based on sensor data; The CNN structure is used to process the generator status data: Input data dimensions: The dimensions of the generator status data are (N, C, H, W), where: N represents the number of samples; C represents the number of channels, which represents the number of different types of sensors; H represents the height of the feature map, which can be the time step or spatial resolution; W represents the width of the feature map, which represents the number of features or the length of the time series Convolutional Layer: The first convolution layer: 64 3x3 convolution kernels, stride 1, using ReLU activation function; The second convolution layer: 128 3x3 convolution kernels, stride 1, using ReLU activation function; The third convolution layer: 256 3x3 convolution kernels, stride 1, using ReLU activation function; Pooling layer: Each convolutional layer is followed by a 2x2 max pooling layer to reduce the size of the feature map; The LSTM structure is used to process the time series data of the remote control mechanism: Input Layer: Input data dimension: The dimension of the remote control mechanism data is (N, L, D), where: N is the number of samples; L is the length of the time series; D is the dimension of the feature, which can be the number of sensors or the number of control parameters Multi-layer LSTM layers: First layer LSTM: number of units: 64, activation function: tanh; Second layer LSTM: number of units: 64, activation function: tanh; The third layer LSTM: number of units: 32, activation function: tanh; Fully connected layer: Fully connected layer 1: Number of neurons: 128, activation function: ReLU; Fusion and fully connected layers: Fusion layer: The outputs of CNN and LSTM can be fused by concatenation operation; Fully connected layer: The first fully connected layer has 1024 neurons and uses the ReLU activation function. The second fully connected layer has 512 neurons and uses the ReLU activation function. In a multi-layer LSTM, each layer of LSTM receives the hidden state of the previous layer as input. The specific process is: The forget gate is used to control the forgetting of part of the information in the memory cell of the previous time step: ; in is the output of the forget gate, is the weight matrix of the forget gate, is the hidden state of the previous time step, is the input of the current time step, is the bias term of the forget gate, σ is the Sigmoid activation function; The input gate controls the amount of new information that enters the memory cell: ; in is the input gate output, is the weight matrix of the input gate, is the hidden state of the previous time step, is the input of the current time step, is the bias term of the input gate, σ is the Sigmoid activation function; The new information is used to update the contents of the memory cell: ; in For new information, To calculate the weight matrix of new information, is the hidden state of the previous time step, is the input of the current time step, is the bias term of the new information, tanh is the hyperbolic tangent function, and is the activation function of the hidden layer; Update memory cells and calculate new memory cell states: ; in is the state of memory cells, is the output of the forget gate, is the memory cell state at the previous time step, is the input gate output, For new information; The output gate controls the information in the hidden state to be passed to the next time step: ; in is the output gate output, is the weight matrix of the output gate, is the hidden state of the previous time step, is the input of the current time step, is the bias term of the output gate; Hide status updates: ; is the hidden state at the current time step, is the output gate output, is the memory cell state at the current time step, tanh is the hyperbolic tangent function, and is the activation function of the hidden layer; Step 3: Use the back propagation algorithm to train the model and optimize the model parameters. The specific process of the back propagation algorithm is to calculate the output of the convolutional neural network CNN. And the output of the long short-term memory network LSTM ; Combine the two to get the fused feature table ; Use the fully connected layer for the final prediction output ; Use the mean square error MSE loss function to measure the prediction results and real results The difference between: ; Back propagation calculates the loss function with respect to the predicted output The gradient is: ; According to the chain rule, the gradient of the loss function is passed back to the fusion layer: ; Pass the gradients of the fusion layer back to each network: ; According to the gradient of each network, the back propagation algorithm is used to update the parameters of each network; according to the gradient calculated by the back propagation algorithm, the parameters of the neural network model are updated using stochastic gradient descent to minimize the loss function; Step 4, divide the data set: divide the data set into training set, validation set and test set; use the status data of the generator and the information of the remote control mechanism as the input of the model, and use the predicted energy consumption as the output of the model; use the back propagation algorithm to train the model and optimize the model parameters so that it can accurately predict the future energy consumption; Depending on the input from different sensors, the following predictions may exist: 1. Temperature: When the internal temperature is between 60℃ and 80℃ and the external temperature is between 20℃ and 40℃, it is predicted that the temperature will remain within this range in the next 6 hours; if the internal temperature suddenly rises to 100℃ and the external temperature rises to 50℃, the model will predict that the temperature will continue to rise in the next 3 hours and remind the operator to check the status of the generator.

[0033] 2. Voltage and current: The voltage is maintained at 220V~240V, and the current is stable at 45A~50A. It is predicted that the power output will remain at a stable level in the next 4 hours. If the voltage suddenly drops to 200V or exceeds 240V and the current rises to 50A, the model will predict that the generator may fail in the next 2 hours and issue an alarm.

[0034] 3. Vibration: The vibration sensor data shows that the vibration amplitude is between 0.1g and 0.2g, and predicts that the vibration level will remain in this range for the next 5 hours; if the vibration amplitude suddenly increases to 0.5g, the model will predict that the vibration will continue to increase in the next 1 hour and recommend maintenance.

[0035] 4. Microphone: Microphone data shows that the noise level is between 50dB and 80dB, and predicts that the sound signal will remain in this range for the next 6 hours; if the noise suddenly increases to 100dB, the model will predict that the noise will continue to increase in the next 30 minutes and recommend maintenance.

[0036] Step 5: Deploy the trained model to the remote visualization control mechanism in the IoT generator system; monitor the status data of the generator and the remote control mechanism information in real time and use them as the input of the model; use the trained model to predict future energy consumption and display the prediction results on the visualization screen to provide timely monitoring and decision support for operators.

[0037] Numerical case analysis: Real-time data: Internal temperature of the generator measured by the temperature sensor: 65°C Generator output voltage measured by voltage sensor: 230V Generator output current measured by current sensor: 20A Vibration level detected by the vibration sensor: Normal Historical load demand data: average load demand is 15kW, peak load demand is 20kW Energy supply data: Powered by diesel engine Remote control organization information: Engine control parameters: normal operating status Load control parameters: The current load demand is stable and no adjustment is required Generator operation mode: automatic operation mode Based on this data, we can predict future energy consumption and display the following on the visualization screen: 1. Current status information: Generator internal temperature: 65°C Generator output voltage: 230V Generator output current: 20A Vibration level: Normal 2. Real-time data display: The real-time status data of the generator is displayed in the form of a chart, including temperature, voltage, current, etc. The current load demand data is updated in real time and displayed on the chart; the power generation and energy consumption of the generator are displayed in real time; 3. Future energy consumption forecast: The prediction results based on the deep neural network model are displayed on the chart, including the energy consumption forecast for a period of time in the future; the predicted energy consumption trend chart is displayed on the visualization screen to help operators understand the changing trend of future energy consumption; 4. Warning and abnormal prompts: If the temperature or voltage is abnormal, the visual screen will display the corresponding warning message and suggest the operator to take appropriate measures; If the vibration level is abnormal, a warning message will also be displayed to alert the operator to possible faults; 5. Maintenance and care reminders: If the generator needs maintenance or service, the visual screen will display the corresponding reminder information and suggest the operator to perform maintenance on time; display the last maintenance time and the recommended next maintenance time; 6. Energy consumption optimization suggestions: If the forecast shows that future energy consumption needs to be optimized, the visualization screen will display the corresponding optimization suggestions and suggest that operators take corresponding measures to reduce energy consumption; Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments, or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An Internet of Things generator with a remote visual control mechanism, comprising a generator body (1), characterized in that: A first generator casing (2) is fixedly mounted on the bottom end of the generator body (1), a second generator casing (3) is provided on the top end of the first generator casing (2), a mounting mechanism (5) is fixedly mounted on one side of the second generator casing (3), the mounting mechanism (5) comprising a mounting base (51) and two height platforms (52), both sides of the top end of the mounting base (51) are respectively fixedly connected to the bottom ends of the two height platforms (52), a synchronization rod (56) is provided at the top end between the two height platforms (52), a visualization screen (11) is fixedly mounted on the middle part of the synchronization rod (56), a remote control mechanism (12) is fixedly mounted on the top end of the visualization screen (11), the remote control mechanism (12) comprises a transmitter (121), a transmitter (122), a controller (123), a receiver (124) and a receiver (125), and the transmitter (121), the transmitter (122), the receiver (124) and the receiver (125) are all connected to the controller (123).

2. The Internet of Things generator with a remote visualization control mechanism according to claim 1, characterized in that: One side of the two height platforms (52) is rotatably connected to a direction rod (55); one end of the two direction rods (55) away from the height platforms (52) is rotatably connected to two ends of a synchronization rod (56); a turntable (53) is installed at the connection between the two height platforms (52) and the two direction rods (55); one side of the two turntables (53) is fixedly installed with a cooling plate (57); one side of the two cooling plates (57) is in contact with and connected to a side directly facing the synchronization rod (56); and a connecting rod (54) is installed between the two turntables (53).

3. The Internet of Things generator with a remote visualization control mechanism according to claim 1, characterized in that: One side of the mounting base (51) is fixedly connected to a side directly opposite to the second generator casing (3).

4. The Internet of Things generator with a remote visual control mechanism according to claim 1, characterized in that: The sending component (122) comprises an input module (1221), a first control module (1222), a first storage module (1223), a recording module (1224), a microphone (1225), an A / D converter (1226) and a wireless signal generating module (1227); the recording module (1224) is connected to the microphone (1225), the microphone (1225) is connected to the A / D converter (1226), and the A / D converter (1226), the wireless signal generating module (1227), the input module (1221) and the first storage module (1223) are all connected to the first control module (1222).

5. The Internet of Things generator with a remote visual control mechanism according to claim 1, characterized in that: The receiving component (125) comprises a playing module (1251), a loudspeaker (1252), a D / A converter (1253), a wireless signal receiving module (1254), a second storage module (1255), a second control module (1256) and a driving module (1257); the playing module (1251) is connected to the loudspeaker (1252); the loudspeaker (1252) is connected to the D / A converter (1253); and the D / A converter (1253), the wireless signal receiving module (1254), the second storage module (1255) and the driving module (1257) are all connected to the second control module (1256).

6. The Internet of Things generator with a remote visual control mechanism according to claim 1, characterized in that: The visualization screen (11) comprises a sensor data display module (111), an input state display module (112), a relay state display module (113), an engine data display module (114), a power generation data display module (115), a load data display module (116) and a fault data display module (117); the sensor data display module (111) displays sensor data in the Internet of Things generator; the input state display module (112) displays the input state of the connector on the Internet of Things generator; the relay state display module (113) displays the relay state on the Internet of Things generator; the engine data display module (114) displays engine data on the Internet of Things generator; the power generation data display module (115) displays power generation data of the Internet of Things generator; the load data display module (116) displays load data of the Internet of Things generator; and the fault data display module (117) displays fault data of the Internet of Things generator.

7. The Internet of Things generator with a remote visual control mechanism according to claim 1, characterized in that: A plurality of first heat dissipation holes (9) are provided on the surface of the first generator casing (2) and the surface of the second generator casing (3); a heat dissipation cover (6) is snap-fitted and mounted on the top end of the second generator casing (3); and a plurality of second heat dissipation holes (10) are provided on the surface of the heat dissipation cover (6).

8. The Internet of Things generator with a remote visual control mechanism according to claim 1, characterized in that: A fixing plate (4) is fixedly mounted on one end of the first generator casing (2).

9. The Internet of Things generator with a remote visual control mechanism according to claim 1, characterized in that: An assembly plate (7) is fixedly mounted on one side of the first generator casing (2) and one side of the second generator casing (3), and a bolt (8) is threadedly connected between the two assembly plates (7).

10. The Internet of Things generator with a remote visual control mechanism according to claim 7, characterized in that: An interface (13) is provided on one side of the heat dissipation cover (6).

11. The state energy consumption prediction method of an Internet of Things generator with a remote visualization control mechanism according to claim 1 is characterized in that: The specific steps include: Step 1: Install sensors to collect real-time status data of the generator and related information transmitted by the remote visualization control mechanism. The sensors used and their installation locations are as follows: Temperature sensor: installed inside the generator body to monitor the operating temperature of the generator; the temperature sensor may be installed on the inner surface of the first generator casing (2) and the second generator casing (3); Voltage sensor and current sensor: installed in the generator output circuit to monitor the voltage and current of the generator; these sensors can be installed at the circuit connection or directly connected to the power line; Vibration sensor: installed on the generator casing surface or bearing to monitor the vibration of the generator; the vibration sensor can help detect whether the generator is operating abnormally; Microphone: used for collecting sound of the recording module (1224) so ​​as to monitor the sound signal during the operation of the generator and to diagnose the operating status of the generator; The required data includes the real-time status data of the generator, including the generator operating status, temperature, voltage, current and other related parameters as well as historical load demand data and energy supply data; Parameter variables: Normal Operation Standby Shutdown Fault Internal Temperature of Generator Generator external temperature (External Temperature) Phase A Voltage Phase B Voltage Phase C Voltage Total Voltage Phase A Current Phase B Current Phase C Current Total Current Average Load Demand Peak Load Demand Load Demand Distribution Electricity Generation Energy Consumption Energy Sources Step 2: Establish a deep neural network model combining a convolutional neural network (CNN) and a long short-term memory (LSTM) network based on sensor data; The CNN structure is used to process the generator status data: Input data dimensions: The dimensions of the generator status data are (N, C, H, W), where: N represents the number of samples; C represents the number of channels, which represents the number of different types of sensors; H represents the height of the feature map, which can be the time step or spatial resolution; W represents the width of the feature map, which represents the number of features or the length of the time series; Convolutional Layer: The first convolution layer: 64 3x3 convolution kernels, stride 1, using ReLU activation function; The second convolution layer: 128 3x3 convolution kernels, stride 1, using ReLU activation function; The third convolution layer: 256 3x3 convolution kernels, stride 1, using ReLU activation function; Pooling layer: Each convolutional layer is followed by a 2x2 max pooling layer to reduce the size of the feature map; The LSTM structure is used to process the time series data of the remote control mechanism: Input Layer: Input data dimensions: The dimensions of the remote control mechanism data are (N, L, D), where: N represents the number of samples; L represents the length of the time series; D represents the dimension of the feature, which can be the number of sensors or the number of control parameters Multi-layer LSTM layers: First layer LSTM: number of units: 64, activation function: tanh; Second layer LSTM: number of units: 64, activation function: tanh; The third layer LSTM: number of units: 32, activation function: tanh; Fully connected layer: Fully connected layer 1: Number of neurons: 128, activation function: ReLU; Fusion and fully connected layers: Fusion layer: The outputs of CNN and LSTM can be fused by splicing operation; Fully connected layer: The first fully connected layer has 1024 neurons and uses the ReLU activation function. The second fully connected layer has 512 neurons and uses the ReLU activation function. In a multi-layer LSTM, each layer of LSTM receives the hidden state of the previous layer as input. The specific process is: The forget gate is used to control the forgetting of part of the information in the memory cell of the previous time step: ； in is the output of the forget gate, is the weight matrix of the forget gate, is the hidden state of the previous time step, is the input of the current time step, is the bias term of the forget gate, σ is the Sigmoid activation function; The input gate controls the amount of new information that enters the memory cell: ； in is the input gate output, is the weight matrix of the input gate, is the bias term of the input gate, σ is the Sigmoid activation function; The new information is used to update the contents of the memory cell: ； in For new information, To calculate the weight matrix of new information, is the bias term of the new information, tanh is the hyperbolic tangent function, and is the activation function of the hidden layer; Update memory cells and calculate new memory cell states: ； in is the state of memory cells, is the output of the forget gate, is the memory cell state at the previous time step, is the input gate output, For new information; The output gate controls the information in the hidden state to be passed to the next time step: ； in is the output gate output, is the weight matrix of the output gate, is the hidden state of the previous time step, is the input of the current time step, is the bias term of the output gate; Hide status updates: ； is the hidden state at the current time step, is the output gate output, is the memory cell state at the current time step, tanh is the hyperbolic tangent function, and is the activation function of the hidden layer; Step 3: Use the back propagation algorithm to train the model and optimize the model parameters. The specific process of the back propagation algorithm is to calculate the output of the convolutional neural network CNN. And the output of the long short-term memory network LSTM ; Combine the two to get the fused feature table ; Use the fully connected layer for the final prediction output ; Use the mean square error MSE loss function to measure the prediction results and real results The difference between: ； Back propagation calculates the loss function with respect to the predicted output The gradient is: ； According to the chain rule, the gradient of the loss function is passed back to the fusion layer: ； Pass the gradients of the fusion layer back to each network: ； According to the gradient of each network, the back propagation algorithm is used to update the parameters of each network; according to the gradient calculated by the back propagation algorithm, the parameters of the neural network model are updated using stochastic gradient descent to minimize the loss function; Step 4, divide the data set: divide the data set into training set, validation set and test set; use the status data of the generator and the information of the remote control mechanism as the input of the model, and use the predicted energy consumption as the output of the model; use the back propagation algorithm to train the model and optimize the model parameters so that it can accurately predict the future energy consumption; Depending on the input from different sensors, the following predictions exist:

1. Temperature sensor data: Forecast under normal temperature conditions: It is predicted that the internal and external temperatures of the generator will remain within the normal range in the next few hours, and the energy consumption will be relatively stable; Prediction of abnormal temperature fluctuations: If the temperature fluctuates abnormally, it indicates that the generator is faulty or the external environment has changed. The model will predict the trend of future temperature fluctuations and provide early warning information; 2. Voltage sensor and current sensor data: Prediction of voltage and current stability: If the voltage and current data remain stable, the model will predict the future operating status and energy consumption of the generator to maintain stable power generation; Prediction of voltage or current anomalies: If the voltage or current is abnormal, it indicates that the generator's output power is unstable or there is a fault. The model will predict the future working status and provide corresponding early warning information; 3. Vibration sensor data: Prediction under normal vibration conditions: predict that the vibration level of the generator will remain within the normal range and the working state will be stable in the future; Prediction of abnormal vibration: If the vibration data shows abnormalities, indicating problems with the generator’s bearings or other components, the model predicts future vibrations and provides fault diagnosis information; 4. Microphone data: Prediction under normal working conditions: predict that the sound signal during the future operation of the generator will remain within the normal range without abnormal conditions; Prediction of abnormal sound: If the microphone data detects abnormal sound, it indicates that the generator is operating abnormally or is faulty. The model will predict the trend of future sound signal changes and provide corresponding early warning information. Step 5: Deploy the trained model to the remote visualization control mechanism in the IoT generator system; monitor the status data of the generator and the remote control mechanism information in real time and use them as the input of the model; use the trained model to predict future energy consumption and display the prediction results on the visualization screen to provide timely monitoring and decision support for operators.

Images