Air pressure type altimeter based on BP neural network

Abstract

The invention provides an air pressure type altimeter based on a BP neural network, and the altimeter comprises a sensor which is used for monitoring atmospheric parameters; the data acquisition module is connected with the sensor and is used for acquiring data containing atmospheric parameters and transmitting the data to the processor module; the processor module is connected with the data acquisition module and is used for carrying out denoising and normalization processing on the data containing the atmospheric parameters; and the upper computer is connected with the processor module, calculates an elevation value according to the trained BP neural network model, and displays the elevation value in real time. The BP neural network can adjust own weight and bias through training so as to adapt to different environment and condition changes, and the adaptive capacity of the altimeter is improved.

Description

Technical Field

[0001] The invention relates to the field of height measuring devices, in particular to a barometric altimeter based on a BP neural network. Background Art

[0002] Traditional barometric altimeters are based on the principle that air pressure changes with altitude. However, the relationship between air pressure and altitude is nonlinear and is easily affected by ambient temperature and humidity. The measurement results of traditional barometric altimeters are easily affected. Summary of the invention

[0003] In view of this, the main purpose of the present invention is to solve the technical problem that the measurement results of the barometric altimeter are easily affected by the ambient temperature and humidity.

[0004] The present invention provides a barometric altimeter based on a BP neural network, comprising:

[0005] A sensor for monitoring an atmospheric parameter;

[0006] A data acquisition module connected to the sensor, the data acquisition module is used to collect data containing the atmospheric parameters and transmit the data to the processor module;

[0007] A processor module connected to the data acquisition module, the processor module being used for performing denoising and normalization processing on the data containing the atmospheric parameters;

[0008] A host computer connected to the processor module calculates the elevation value according to the trained BP neural network model and displays the elevation value in real time.

[0009] In some embodiments of the present invention, the barometric altimeter further includes a power supply module, which is respectively connected to the sensor, the data acquisition module and the processor module for supplying power.

[0010] In some embodiments of the present invention, the sensor includes a temperature sensor, an air pressure sensor, and a humidity sensor, which are used to monitor the ambient temperature, the atmospheric pressure, and the ambient humidity, respectively.

[0011] In some embodiments of the present invention, the BP neural network model calculation process is:

[0012] Forward propagation:

[0013] The input layer consists of 4 neurons, representing atmospheric pressure, ambient temperature, ambient humidity and time respectively;

[0014] Output of the input layer: x i(i=1, 2, 3, 4);

[0015] The hidden layer consists of 10 neurons, and the weights of the input layer and the hidden layer are: ki (i=1, 2, 3, 4; k=1, 2, 3....10);

[0016] Hidden layer input:

[0017] Hidden layer output: z k =f(S k );

[0018] Activation function, use sigmoid activation function;

[0019] The output layer consists of 1 neuron, which represents the elevation value. The output layer input is:

[0020] Output layer output: y = f(S j );

[0021] Backward Propagation:

[0022] If the actual output of the output layer does not match the expected output, back propagation begins;

[0023] The weights of the hidden layer and the output layer are:

[0024]

[0025] The two weights have a common part and can be simplified to:

[0026]

[0027] In the formula, t j Indicates the expected output; y j represents the actual output; η represents the learning rate; z k represents the output of the hidden layer; x i represents the input of the input layer; w jk represents the weights of the hidden layer and the output layer; s k and j Represent the input of the output layer and the hidden layer respectively; f′ z and f′ are the activation functions of the hidden layer and the output layer;

[0028] Weight update:

[0029] The correction and update of the neural network mainly involves updating and correcting the weights. The update formula is:

[0030]

[0031] wij ←w ij +Δw ij .

[0032] In some embodiments of the present invention, the data acquisition module and the processor module are connected via I 2 C communication connection.

[0033] The present invention provides a barometric altimeter based on BP neural network, which has the following beneficial effects:

[0034] Use a combination of multiple sensors, use pressure sensors to obtain accurate atmospheric pressure values, use temperature sensors and humidity sensors to obtain temperature and humidity data to compensate for atmospheric pressure, and improve the accuracy of calculated elevation data;

[0035] The BP neural network model is used to calculate the elevation value, and the good nonlinear mapping ability of the BP neural network is used to effectively capture the nonlinear relationship between the input parameters (air pressure, temperature, humidity and time) and the elevation;

[0036] The BP neural network can adjust its own weights and biases through training to adapt to different environments and conditions, thus improving the adaptability of the altimeter.

[0037] Each part of the system is connected to the MCU through a standard communication interface. Both hardware design and software programming adopt a modular structure to facilitate system maintenance and upgrades. Through data denoising and normalization processing, the reliability and stability of the system in various environments are ensured;

[0038] The system does not require the input of additional parameters and is easy to operate and maintain. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 An electrical principle block diagram of an embodiment of the present invention;

[0040] Figure 2 Schematic diagram of a neural network learning process according to an embodiment of the present invention. DETAILED DESCRIPTION

[0041] The present application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It will be appreciated that the specific embodiments described herein are only used to explain the relevant inventions, rather than to limit the inventions. It should also be noted that, for ease of description, only portions related to the relevant inventions are shown in the accompanying drawings. It should be noted that, in the absence of conflict, the embodiments in the present application and the features in the embodiments may be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

[0042] See also Figure 1,A barometric altimeter based on BP neural network is mainly composed of a temperature sensor, a barometric sensor, a humidity sensor, a data acquisition module, a processor module, a power module and a host computer software.

[0043] Specifically, sensors,are used to monitor atmospheric parameters;

[0044] A data acquisition module connected to the sensor, the data acquisition module is used to collect data including atmospheric parameters and transmit them to the processor module;

[0045] A processor module connected to the data acquisition module, the processor module is used to perform denoising and normalization processing on the data containing the atmospheric parameters;

[0046] The host computer connected to the processor module calculates the elevation value according to the trained BP neural network model and displays the elevation value in real time.

[0047] Please note that Figure 2 , the calculation process of BP neural network model is:

[0048] Forward propagation:

[0049] The input layer consists of 4 neurons, representing atmospheric pressure, ambient temperature, ambient humidity and time respectively;

[0050] Output of the input layer: x i (i=1, 2, 3, 4);

[0051] The hidden layer consists of 10 neurons, and the weights of the input layer and the hidden layer are: ki (i=1, 2, 3, 4; k=1, 2, 3....10);

[0052] Hidden layer input:

[0053] Hidden layer output: z k =f(S k );

[0054] Activation function, use sigmoid activation function;

[0055] The output layer consists of 1 neuron, which represents the elevation value. The output layer input is:

[0056] Output layer output: y = f(S j );

[0057] Backward Propagation:

[0058] If the actual output of the output layer does not match the expected output, back propagation begins;

[0059] The weights of the hidden layer and the output layer are:

[0060]

[0061] The two weights have a common part and can be simplified to:

[0062]

[0063] In the formula, t j Indicates the expected output; y j represents the actual output; η represents the learning rate; z k represents the output of the hidden layer; x i represents the input of the input layer; w jk represents the weights of the hidden layer and the output layer; s k和 s j Represent the input of the output layer and the hidden layer respectively; f′ z and f′ are the activation functions of the hidden layer and the output layer;

[0064] Weight update:

[0065] The correction and update of the neural network mainly involves updating and correcting the weights. The update formula is:

[0066]

[0067] w ij ←w ij +Δw ij .

[0068] The working principle is: the sensor is used to monitor the atmospheric pressure, ambient temperature and ambient humidity in real time. A variety of sensors are used in combination. The pressure sensor is used to obtain accurate atmospheric pressure values. The temperature sensor and humidity sensor are used to obtain temperature and humidity data to compensate for the atmospheric pressure. The host computer software obtains the current time and combines multiple parameters that affect the elevation to improve the accuracy of the calculated elevation data. The power module provides power for the entire system. The data acquisition module transmits the sensor data to the processor module via O2C communication. The processor module denoises and normalizes the collected data. The processor module sends the pre-processed data to the PC through serial port communication, and uses the trained BP neural network model to calculate the elevation value. The good nonlinear mapping ability of the BP neural network is used to effectively improve the adaptability of the barometric altimeter in complex environments and ensure the accuracy of the results. Finally, the elevation data obtained is displayed in real time through the host computer software.

[0069] The above are only some embodiments of the present invention. For those skilled in the art, several modifications and improvements can be made without departing from the creative concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. A barometric altimeter based on BP neural network, characterized in that: include: A sensor for monitoring an atmospheric parameter; A data acquisition module connected to the sensor, the data acquisition module is used to collect data containing the atmospheric parameters and transmit the data to the processor module; A processor module connected to the data acquisition module, the processor module being used for performing denoising and normalization processing on the data containing the atmospheric parameters; A host computer connected to the processor module calculates the elevation value according to the trained BP neural network model and displays the elevation value in real time.

2. A barometric altimeter based on BP neural network according to claim 1, characterized in that: The barometric altimeter further comprises a power supply module, which is respectively connected to the sensor, the data acquisition module and the processor module for supplying power.

3. The barometric altimeter based on BP neural network according to claim 1, characterized in that: The sensors include a temperature sensor, an air pressure sensor and a humidity sensor, which are used to monitor the ambient temperature, the atmospheric pressure and the ambient humidity respectively.

4. A barometric altimeter based on BP neural network according to claim 1, characterized in that: The BP neural network model calculation process is: Forward propagation: The input layer consists of 4 neurons, representing atmospheric pressure, ambient temperature, ambient humidity and time respectively; Output of the input layer: x i (i=1,2,3,4); The hidden layer consists of 10 neurons, and the weights of the input layer and the hidden layer are: ki (i=1, 2, 3, 4; k=1, 2, 3....10); Hidden layer input: Hidden layer output: z k =f(S k ); Activation function, use sigmoid activation function; The output layer consists of 1 neuron, which represents the elevation value. The output layer input is: Output layer output: y = f(S j ); Backward Propagation: If the actual output of the output layer does not match the expected output, back propagation begins; The weights of the hidden layer and the output layer are: The two weights have a common part and can be simplified to: Where, t j Indicates the expected output; y j represents the actual output; η represents the learning rate; z k represents the output of the hidden layer; x i represents the input of the input layer; w jk represents the weights of the hidden layer and the output layer; s k and j Represent the input of the output layer and the hidden layer respectively; f′ z and f′ are the activation functions of the hidden layer and the output layer; Weight update: The correction and update of the neural network mainly involves updating and correcting the weights. The update formula is: In ij ←in ij +Δw ij 。 5. The barometric altimeter based on BP neural network according to claim 1, characterized in that: The data acquisition module and the processor module are connected via I 2 C communication connection.

Images