Internet of things array receiver for power transient electromagnetic instrument

Abstract

The application discloses an internet of things array receiver of a power transient electromagnetic instrument, and the receiver structure is a two-section design of a host computer and a receiving coil, and the host computer adopts a modular circuit board laminated design. The receiving coil comprises a connecting head connected through a connecting line and a coil receiving antenna printed on a circuit board. The host computer comprises a quick socket, a power circuit, an analog signal processing circuit, a digital control circuit, a switch and an aviation socket which are packaged in a shell. The connecting head is connected with the quick socket. The application can be used in combination with the transient electromagnetic instrument on the market, and rich interface transmission data are reserved. The receiver solves the problems of limited application scene and complicated data acquisition of the transient electromagnetic instrument on the market. In traditional application, the operation flexibility and efficiency can be improved, in multi-channel array application, the human error transmission rate can be greatly reduced, and the data processing efficiency is improved through the construction of a network database.

Description

Technical Field

[0001] This invention relates to the field of geology and exploration geophysics, and in particular to an Internet of Things (IoT) array receiver for a power-dependent transient electromagnetic instrument. Background Technology

[0002] Currently, geological exploration primarily relies on drilling technology, especially in projects like mine and tunnel construction, where extensive drilling data is essential for determining construction plans. This method is time-consuming, labor-intensive, and costly. Meanwhile, in recent years, some non-contact exploration methods, such as simulated seismic wave methods and electromagnetic inversion methods, have gradually moved from theoretical application to practical engineering. However, these methods have encountered significant resistance in some applications. Transient electromagnetic methods (TEMs), a type of electromagnetic inversion method, are frequently used in geological research. To date, numerous companies and universities have developed various TEM instruments. The basic principle of these instruments is to pass a pulsed current through a loop coil to excite a "primary electromagnetic field." As this "primary field" diffuses underground, it generates a "secondary induced field" through electromagnetic induction in different substances. By collecting and processing this "secondary field," the resistivity structure of the underground can theoretically be analyzed, thereby revealing the geological structure. Extensive research and experimentation have proven the feasibility of this method, and it is hoped that this technology can significantly reduce drilling operations and save human and material resources. However, the research conditions for TEMs are demanding, the equipment development cycle is long, and it targets a small group of people, resulting in high costs and hindering its widespread application in various scenarios.

[0003] The limitations of transient electromagnetic methods in both research and application stem partly from the fact that the transmitters and receivers of transient electromagnetic instruments remain relatively primitive, often employing a bonded structure that requires them to be used together. This results in limited adjustable parameters and poor consistency between devices. Furthermore, most of these devices lack automation and portability, making them unsuitable for use in confined spaces. Multichannel transient electromagnetic instruments also suffer from significant on-site installation work, inaccurate data recording, and time-consuming, labor-intensive, and error-prone inversion processes. Additionally, processing transient electromagnetic data requires advanced professional knowledge, leading to high training costs. In summary, complex construction and cumbersome operation are pressing issues that need to be addressed in the application of electromagnetic methods. Summary of the Invention

[0004] The purpose of this invention is to solve the complex construction and post-processing problems of current transient electromagnetic detection projects, and to provide an IoT array receiver for a power-operated transient electromagnetic instrument. This invention can perform multi-channel synchronous transient electromagnetic directional detection, and is easy to install and construct, with rich functions.

[0005] The structure of the IoT array receiver for the power transient electromagnetic instrument includes a host and a receiving coil. The receiving coil includes a coil receiving antenna printed on a circuit board and a connector connected by a connecting wire. The host includes a quick-connect socket, a battery pack, a multi-rail Buck-Boost circuit, an analog signal processing circuit, a digital control circuit, a switch, and an aviation socket, all encapsulated in a housing. The connector and the quick-connect socket are connected accordingly. The quick-connect socket is connected to the analog signal processing circuit. The analog signal output terminal of the analog signal processing circuit is connected to the digital control circuit. The filtering module in the analog signal processing circuit is connected to the coaxial interface in the aviation socket. The external port of the digital control circuit is connected to the aviation socket. The power circuit includes a battery pack and a multi-rail Buck-Boost circuit. The power terminals of the analog signal processing circuit and the digital control circuit are connected to the power output terminal of the multi-rail Buck-Boost circuit. The charging management chip in the multi-rail Buck-Boost circuit is connected to the switch and the charging interface in the aviation socket.

[0006] Multiple IoT array receivers as described in this invention are placed in different locations and then synchronously receive signals from a signal source from different angles.

[0007] Specifically, the aviation socket includes: a coaxial interface, a differential interface, a USB interface, a UART interface, an RG45 interface, a function enable interface, and a charging interface. The coaxial interface serves as the output terminal for transient electromagnetic analog signals, and the differential interface serves as the input for system expansion synchronization signals. The USB interface is used to connect other receiver hosts and computers. The UART interface is used to transmit processed transient electromagnetic digital signals. The RG45 interface is used to connect the receiver to a wired network. The function enable interface is used for external configuration of the receiver's function mode. The charging interface is used to power and charge the receiver.

[0008] Specifically, the host adopts a modular circuit board stacked design, with the circuit boards containing the battery pack, multi-rail Buck-Boost circuit, analog signal processing circuit, and digital control circuit stacked one on top of the other, and the quick-connect socket, switch, and aviation socket set on the outer shell.

[0009] Specifically, the digital control circuit includes: a Cortex-A35 processor and an FPGA coprocessor interconnected with each other; the FPGA coprocessor is connected to the digital signal output of the analog signal processing circuit, the IoT module, the Zigbee chip, and the attitude sensor; the Cortex-A35 processor is connected to a memory chip, a RAM chip, an RG45 network signal transformer, and an Ethernet PHY chip; the Cortex-A35 processor is used to run an operating system to realize module communication, resource scheduling, and data processing; the FPGA coprocessor is used to bridge the Cortex-A35 processor, the IoT module, the Zigbee chip, the attitude sensor, and the analog signal processing circuit; the IoT module is used to access wireless... The system includes internet connectivity, location tracking, and clock calibration. The Zigbee chip is controlled by an FPGA coprocessor to build a local self-organizing network, forming a receiver array that works with the IoT module for precise positioning and data exchange. The attitude sensor assists in positioning and determines the exploration direction when receiving signals. The memory chip stores the system program. The memory chip is an external extended memory space for the Cortex-A35. The RG45 network signal transformer and Ethernet PHY chip bridge the RG45 interface network and the Cortex-A35, enabling wired network connectivity. The analog signal processing circuit includes a precision instrument amplifier, a filter circuit, and a high-speed analog-to-digital converter connected in sequence; the precision instrument amplifier is connected to a quick-connect socket.

[0010] The filtering circuit described in this invention employs an adaptive mean and Gaussian hybrid filtering algorithm to filter the transient electromagnetic signal sequence A′(m,n)=(m,n)+I(m,n)+T(m,n), where (m,n) represents the nth sample value of the mth superposition sampling, I(m,n) represents the instrument intermodulation interference noise, and T(m,n) represents the instrument hot electron spurious noise. The steps are as follows:

[0011] (1) Sampling: Divide the signal sequence A′(1,n) of the target row l into segments of equal length according to the time interval, and then extract a segment of data to be processed A″(n). Since the length of the A″(n) sequence is very short, it can be regarded as an additive function of a linear function and white noise.

[0012] (2) Noise Prediction: The first moment E, second moment D, slope K of the linear function fitting, energy spectral density P, and power spectral density S of A″(n) are calculated. The sum of these basic parameters and a random number is used as the basis for prediction. Then, the baseline of A″(n) is corrected using the prediction basis to predict the first moment. The sequence A″′(n) is used, where N represents the sequence length, and this sequence represents the noise probability characteristics in the A″(n) sequence. Then, the joint distribution and independence of A″(n) and A″′(n) are calculated to verify the reliability of the prediction. If it is unreliable, a new random number is generated and added to the previous basic parameters, and the prediction is repeated until the prediction is reliable. After that, each row of A′(m,n) is sampled and noise prediction is performed to finally obtain A″(n) and A″′(n).

[0013] (3) Estimating the likelihood: Calculate the convolution kernel based on the energy ratio of A″(n) to A″′(n), the first moment of A″′(n), and the second moment of A″′(n) in each superimposed sampling sequence. The likelihood values ​​of the three parameters a, b, and c are recalculated based on the obtained values ​​of a, b, and c each time.

[0014] (4) Calculate the convolution: Take the average of A′(1,n) to obtain Then, convolve the result with the convolution kernel f(n) to obtain the filtered result;

[0015] (5) Repeat steps (1) to (4) to complete the noise processing for each row of A′(m,n) to complete the filtering.

[0016] Specifically, the Zigbee chip uses the Zigbee simple chain protocol. The protocol's data packet structure includes nine data areas: MAC-R, MAC-T, REQ / ACK, State, Frame, DS, Data, DE, and CRC. These represent the target device's MAC address, the current device's MAC address, the response status, the device status, the data frame, the start of data flag, the data content, the end of data flag, and the checksum, respectively. When the REQ / ACK data bit is 0, it represents REQ: requesting the target device to communicate; when it is 1, it represents ACK: responding to the target device's request. State represents several device states: (00, Ready) can receive data, (01, Busy) busy and unable to receive data, (10, Finish) data received successfully, and (11, Retrans) data received successfully but incompletely, requesting retransmission.

[0017] The communication process of the Zigbee shortchain protocol includes:

[0018] (1) Device A sends a REQ request to Device B with State = 00. If no response is received from Device B after the set time, Device A will resend the request.

[0019] (2) When device B receives the request, it will immediately send an ACK response to device A according to its own status. If device B is busy, it will return State=01 and will not respond to device A's request for a set time. Otherwise, it will return State=00 and then establish communication with device A.

[0020] (3) When device A receives a status response from device B and State = 00, device A will send a REQ request containing data to device B; if device B does not respond within the set time, device A will resend the request.

[0021] (4) When device B receives data from device A, it first verifies the data and then sends the result to device A via ACK response. If the data verification fails, the state bit State = 11. At this time, device A will repeat step 3. Otherwise, State = 10, and one round of requests ends.

[0022] The process of constructing a network using Zigbee among multiple receivers of this invention and optimizing the network using a neuromorphic network optimization algorithm includes:

[0023] (1) Each receiver simulates the growth of neuronal synapses, sends connection requests to the outside and receives responses through the Zigbee chip, and records the other party's MAC address and signal strength.

[0024] (2) Through several rounds of cycles and information exchange, the receiver establishes a local area network dictionary and initially constructs an unordered network through the dictionary; the dictionary records the MAC code, connectivity, signal strength, and number of connections of other receivers in the area; the lines represent receiver links that can communicate, and the numbers represent the number of links that the receiver can establish.

[0025] (3) The receiver optimizes its own links based on the number of links of the surrounding receivers and builds a routing table to record information and pass the links; the receivers at the edge will start the optimization first, prioritizing the retention of links with receivers with more links, removing paths with other receivers, and then passing these changes to other receivers to update the local area network dictionary; after several rounds of retention and removal processes, the edge receiver will have only one link left, which will be retained in the routing table and passed to other receivers, while these links will no longer participate in the optimization process; the receivers at the center will optimize in the same way until all links in the network are in the routing table;

[0026] (4) When a new receiver joins an existing network, it broadcasts a connection request to receivers in the network and assigns it to a node or coordinator. If the receiver is close to an existing node, it will connect to that node. If there are no nodes around but it is close to one of the terminals, it will send a connection request to that terminal, and then that terminal becomes a node and connects the receiver to the network.

[0027] The beneficial effects of this invention are: it can be used in conjunction with commercially available transient electromagnetic instruments and provides abundant interfaces for data transmission. This receiver solves the problems of limited application scenarios and cumbersome data acquisition associated with current commercially available transient electromagnetic instruments. In traditional applications, it improves operational flexibility and efficiency; in multi-channel array applications, it significantly reduces the rate of human error in data transmission; and by constructing a network database, it improves data processing efficiency. Attached Figure Description

[0028] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0029] Figure 1 This is a schematic diagram of the design structure of the present invention.

[0030] Figure 2 This is a circuit block diagram of the present invention.

[0031] Figure 3 This is the receiving coil board of the present invention.

[0032] Figure 4 This is a set of noisy signal sequences in the example.

[0033] Figure 5 for Figure 4 The result of filtering the noise sequence using an adaptive mean and Gaussian mixture filter.

[0034] Figure 6 This is a data packet for the Zigbee simplified chain protocol of this invention.

[0035] Figure 7 This is a schematic diagram of the response process of the Zigbee simple chain protocol of the present invention.

[0036] Figure 8 This paper describes the process of establishing and optimizing a Zigbee network using a neural biomimetic network optimization algorithm. Detailed Implementation

[0037] The structure and working principle of the IoT array receiver of the power-dependent transient electromagnetic instrument of the present invention will be described in detail below with reference to the accompanying drawings.

[0038] like Figure 1 , Figure 2As shown, the IoT array receiver structure of the power-dependent transient electromagnetic instrument of the present invention is a two-stage design of host and receiving coil. The host is rectangular in shape and adopts a modular circuit board stacked design. The receiving coil includes a coil receiving antenna 1-1 (referred to as coil board) printed on the circuit board and a connector 1-3 connected by a connecting wire 1-2. The coil board 1-1 adopts a multi-turn small coil structure with a center tap and an internal magnetic core, which is small in size and has high induction sensitivity. The coil board 1-1 is manufactured by PCB board fabrication. While receiving signals, it can effectively collect weak signals due to the anti-metal interference design. The induction coil has three wires leading out and is soldered to the three spring terminals of the connector 1-3 by wires.

[0039] The so-called array reception refers to placing multiple identical receivers of the present invention in different positions, and then synchronously receiving a signal source from different angles.

[0040] The main unit includes a quick-connect socket 3, a battery pack 4, a multi-rail Buck-Boost circuit 5, an analog signal processing circuit 6, a digital control circuit 7, a switch 8, and an aviation socket 9, all encapsulated in a housing 2. Connectors 1-3 are connected to the quick-connect socket 3, which in turn connects to the analog signal processing circuit 6. The analog signal output of the analog signal processing circuit 6 is connected to the digital control circuit 7. The filtering module in the analog signal processing circuit 6 is connected to the coaxial interface in the aviation socket 9. The external port of the digital control circuit 7 is connected to the aviation socket 9. The power circuit includes the battery pack 4 and the multi-rail Buck-Boost circuit 5. The power terminals of the analog signal processing circuit 6 and the digital control circuit 7 are connected to the power output of the multi-rail Buck-Boost circuit 5. The charging management chip in the multi-rail Buck-Boost circuit 5 is connected to the switch 8 and the charging interface in the aviation socket 9. The main unit is capable of analog signal processing, digital signal processing, positioning, internet access, and networking functions.

[0041] The main unit's power supply circuit consists of a battery pack 4 and a multi-rail Buck-Boost circuit 5. The battery pack uses six lithium manganese oxide batteries connected in series to power the device. The maximum voltage is 25.2V, the minimum voltage is 17V, and the rated current is 1A, meeting the explosion-proof requirements for hazardous environments. Figure 2The multi-rail Buck-Boost circuit 5 includes several voltage regulator circuits such as BQ24650, LTC3892, MP9442, TPS7A49, and TPS7A30, which perform DC / DC boost, buck, and regulation functions respectively, generating the required ±16V, ±12V, 5V, 3V, 1.8V, 1.5V, and 1.2V multi-rail voltages. The BQ24650 is the input to the battery pack, used for charging and discharging management of the lithium battery pack. The LTC3892 draws power from the battery pack to generate the ±16V rail, which then passes through the TPS7A49 and TPS7A30 adjustable linear regulators to obtain the ±12V rail required by the operational amplifier circuit. The MP9442 draws power from the battery pack to generate the 5V, 3.3V, 1.8V, 1.5V, and 1.2V rails required by the digital control circuit 7. The rail outputs of TPS7A49, TPS7A30, and MP9442 are connected to the analog signal processing circuit 6 and the digital control circuit 7 respectively via the power bus.

[0042] The host's analog signal processing circuit 6 includes a precision instrument amplifier, a filter circuit, and a high-speed analog-to-digital converter (14-bit high-speed A / D). The input of the precision instrument amplifier is connected to the induction coil via quick-connect socket 3, and its output is connected to the input of the filter module. The output of the filter module is connected to the analog input of the 14-bit high-speed A / D. The precision instrument amplifier amplifies the induced voltage signal from the induction coil, achieving preliminary amplification of the small signal. The filter module consists of four ADA3894s, designed as a 5th-order Sallen-key analog bandpass filter circuit with a bandpass of 200Hz to 600kHz, used to filter out power frequency interference in the signal. The 14-bit high-speed A / D converts the amplified and filtered electromagnetic induction analog electrical signal into a 14-bit precision digital signal, which is then transmitted to the digital control circuit for further processing via the data bus.

[0043] The digital control circuit 7 includes an interconnected Cortex-A35 processor and an FPGA coprocessor. The FPGA coprocessor is connected to the digital signal output of the analog signal processing circuit 6, the IoT module, the Zigbee chip, and the attitude sensor. The Cortex-A35 processor is connected to a memory chip, a RAM chip, an RG45 network signal transformer, and an Ethernet PHY chip. In this embodiment, the Cortex-A35 processor is an RK3326, which includes several serial ports and programmable GPIO interfaces for running a Unix system, enabling module communication, resource scheduling, and data processing. The FPGA coprocessor is a Spartan7-XC7S15, used to bridge the Cortex-A35 processor, IoT module, Zigbee chip, attitude sensor, and 14-bit high-speed A / D converter. The IoT module is an RM510Q-GL, supporting 4G / 5G mobile networks and GPS positioning, used for accessing the wireless internet, positioning, and clock calibration. The Zigbee chip uses an RFX2401C front-end modem, controlled by an FPGA coprocessor, to build a local self-organizing network, forming a receiver array, and working with the IoT module for precise positioning and data exchange. The attitude sensor is an MPU9250, used to assist positioning and determine the exploration direction when receiving signals. Since exploration operations are directional—for example, when exploring groundwater, the receiver is held vertically towards the ground; when probing a dam, the receiver needs to be inserted into a slanted borehole—without an attitude sensor, the exploration direction cannot be determined. The memory chip stores the system program. The DDR4 memory chip is an external extended memory space for the Cortex-A35. The RG45 network signal transformer and Ethernet PHY chip bridge the RG45 interface network and the Cortex-A35, enabling wired network connectivity. Both the Zigbee and IoT modules are connected to their respective RF antennas for wireless network signal transmission and reception.

[0044] The Cortex-A35 processor is connected to the memory, DDR4 memory, and RG45 Ethernet via a bus to form a microcomputer system running the Unix operating system. The FPGA coprocessor, leveraging its parallel computing, rich I / O, and programmable features, connects the IoT module, Zigbee circuitry, attitude sensor, and 14-bit A / D converter to the FPGA's internal bus via a data bus. It builds an asynchronous buffer pool between different components to achieve efficient data exchange, and then connects to the Cortex-A35 processor via the data bus to perform coprocessing functions.

[0045] The external connectors of the main unit include quick-connect socket 3, push-button switch 8, and aviation socket 9. Quick-connect socket 3 is paired with connector 1-3 to connect the three terminals of the induction coil and the input terminal of the precision instrument amplifier. Connector 1-3 is connected to the coil board, and quick-connect socket 3 is connected to the input terminal of the precision instrument amplifier.

[0046] The push-button switch 8 is connected to the enable pins of the LTC3892 and MP9442 chips on the power supply circuit. By controlling the startup of the power supply chip, the power supply circuit is controlled to turn the system on and off. The three status lights integrated on the push-button switch 8, namely red, yellow and green, represent charging, network and working indicators, respectively.

[0047] The aviation socket 9 uses a 28-pin LEMO connector, including a coaxial interface (dual-pin), a differential interface (dual-pin), a USB 2.0 interface (4-pin), a UART interface (4-pin), an RG45 interface (8-pin), a function enable interface (6-pin), and a charging interface (dual-pin). The coaxial interface serves as the output terminal for transient electromagnetic analog signals, and the differential interface serves as the input for system expansion synchronization signals. The USB 2.0 interface is used to connect the receiver host and the PC; the UART interface is used to transmit processed transient electromagnetic digital signals; the RG45 interface is used to connect the receiver to a wired network; the function enable interface is used for external configuration of the receiver's function modes; and the charging interface is used to power and charge the receiver. The charging interface connects to the BQ24650 input terminal of the power supply circuit; the function enable interface connects to the jumper terminal on the digital control circuit to independently control the operating state of the component; the USB interface and UART interface connect to the Cortex-A35 processor; the RG45 interface connects to the RG45 Ethernet component; the differential line connects to the FPGA via a voltage protector for external synchronization signal input; the coaxial line connects to the output of the filter module in the analog signal processing circuit for direct analog signal output.

[0048] The construction method will be flexibly adjusted according to the object being explored. For large-area exploration in unknown environments, hundreds of receivers can be arranged in a square array spaced 5 to 20 meters apart to achieve carpet-like detection; if the object being explored is a serpentine vein, the receivers can be arranged along the vein; for vertical shafts, the receivers can be placed in boreholes inside the shaft. In these applications, the receivers will be located via Zigbee network and GPS, attitude sensors, and then the relevant data will be saved or uploaded to a network database for unified management by a host computer, eliminating the need for manual on-site data processing.

[0049] Furthermore, since the receiver has a built-in battery, it can be deployed simply by charging the device, and no cables need to be connected during operation.

[0050] The key algorithms used in this invention are described below.

[0051] This invention employs an adaptive mean and Gaussian hybrid filtering algorithm, which is a hybrid filter based on a maximum-minimum digital filter and a Gaussian white noise filter, used to denoise and smooth the acquired transient electromagnetic signals. Figure 4 , Figure 5 This is the effect of an adaptive mean and Gaussian mixture filter.

[0052] In this invention, the adaptive mean and Gaussian mixture filtering algorithm is used to smooth transient electromagnetic signals. The reason for developing this algorithm is that a raw transient electromagnetic signal sequence can be expressed as...

[0053] A(m,n) = B(m,n) + xH(m,n) + yG(m,n) + zI(m,n) + rT(m,n), where B(m,n) represents the effective transient electromagnetic signal, which is an exponentially decaying time-domain signal; H(m,n), G(m,n), I(m,n), and T(m,n) represent natural electromagnetic noise, artificial electromagnetic noise (or power frequency interference), instrument intermodulation interference noise, and instrument hot electron stray noise, respectively; and x, y, z, and r are the noise intensity coefficients. These signal / noise sequences are all matrices, where (m,n) represents the nth sample value from the m-th superposition sampling. After A(m,n) passes through the filtering circuit of the analog signal processing circuit, most of the power frequency interference A(m,n) is outside the design passband of the filtering circuit and will be filtered out. However, natural electromagnetic noise H(m,n), instrument intermodulation interference noise I(m,n), and instrument hot electron stray noise T(m,n) often exhibit Gaussian white noise characteristics and cannot be effectively filtered by analog filters of finite order. Therefore, digital filters are required. Mean filters, commonly used to eliminate white noise, easily cause loss of detail in transient electromagnetic signals. Therefore, a dynamic filter specifically designed for transient electromagnetic signals is needed.

[0054] The adaptive mean and Gaussian mixture filtering algorithm has four main steps for filtering the transient electromagnetic signal sequence A′(m,n)=(m,n)+I(m,n)+T(m,n) after passing through the filtering circuit: sampling, noise prediction, likelihood estimation, and convolution.

[0055] (1) The sampling process will divide the signal sequence A′(1,n) of the target row l into segments of equal length according to the time interval, and then extract a segment of data to be processed A″(n). Since the length of the A″(n) sequence is very short, it can be regarded as an additive function of a linear function and white noise.

[0056] (2) In the noise prediction stage, the first moment E, second moment D, slope K of the linear function fitting, energy spectral density P, and power spectral density S of A″(n) are calculated. The result of adding these basic parameters to a random number is used as the prediction basis. Then, the baseline of A″(n) is corrected using the prediction basis to predict the first moment. The sequence A″′(n) is used, where N represents the sequence length, and this sequence represents the noise probability characteristics in the A″(n) sequence. Then, the joint distribution and independence of A″(n) and A″′(n) are calculated to verify the reliability of the prediction. If it is unreliable, a new random number is generated and added to the previous basic parameters, and the prediction is repeated until the prediction is reliable. After that, each row of A′(m,n) is sampled and noise is predicted, and finally A″(n) and A″′(n) are obtained.

[0057] (3) In the likelihood estimation stage, the energy ratio of A″(n) to A″′(n), the first moment of A″′(n), and the second moment of A″′(n) in each superimposed sampling sequence will be used to calculate the convolution kernel. The three parameters are a, b, and c.

[0058] A″′(n) first moment

[0059] A″′(n) second moment

[0060] Energy ratio

[0061] Solve the above three equations simultaneously to find a, b, and c.

[0062] The likelihood value is recalculated based on the values ​​of a, b, and c obtained each time.

[0063] The formula for calculating the likelihood value is: The likelihood values ​​for b and c are similar.

[0064] (4) In the convolution step, the average of A′(1,n) is calculated to obtain Then, convolve it with the convolution kernel f(n) to obtain the filtered result.

[0065] Repeat the above four steps to complete the noise processing of each row of A′(m,n) to complete the filtering. Figure 3 A set of noisy signals; Figure 4 To apply this filter pair Figure 3 The result of filtering the signal.

[0066] The networking method of this invention adopts a Zigbee Simple Link Protocol, which is based on the 802.15.4 physical protocol layer. This protocol is specifically designed for the communication requirements of this invention, and optimizes the session structure, logic, and data structure complexity of messages to establish a simple and efficient Zigbee wireless network link.

[0067] Figure 6 The data packet structure of the Zigbee simple chain protocol is used in this invention.

[0068] The Zigbee simplified chain protocol defines a minimal protocol structure based on the 802.15.4 physical protocol layer to meet the application requirements of this invention. This protocol adopts a non-beacon mode, allowing Zigbee devices to operate continuously. Although this increases power consumption, it enables them to complete a request within 2ms. The protocol's data packet structure includes nine data areas: MAC-R, MAC-T, REQ / ACK, State, Frame, DS, Data, DE, and CRC. These represent the target device's MAC address, the current device's MAC address, the response status, the device status, the data frame, the start of data flag, the data content, the end of data flag, and the checksum, respectively. In the data packet, a REQ / ACK bit of 0 represents REQ: a request for communication from the target device; a REQ bit of 1 represents ACK: a response to the target device's request. The State data packet represents several device states: (00, Ready) can receive data; (01, Busy) busy, unable to receive data; (10, Finish) data received successfully; (11, Retrans) data received successfully but incompletely, requesting retransmission. The Data Payload is the effective data transmitted in the Zigbee network. This data is typically parallel, short, and discrete in structure, and needs to be directly processed and parsed by the FPGA coprocessor. Therefore, direct transmission is more efficient than compression-transmission-decompression. The Data Payload is also defined to contain only 19 symbols: 0-9, <, >, +, -, $, forward slash, colon, dot, and space. It uses 5-bit encoding instead of 8-bit ASCII, reducing the spatial complexity of the data and further improving transmission speed and efficiency.

[0069] Figure 7 The four-step communication process of the Zigbee simplified chain protocol is demonstrated.

[0070] First, device A sends a REQ request to device B with State = 00. If no response is received from device B within 20ms, device A will resend the request.

[0071] When device B receives the request, it will immediately send an ACK response to device A based on its own status. If device B is busy, it will return State=01 and will not respond to device A's request for 100 milliseconds; otherwise, it will return State=00 and then establish communication with device A.

[0072] When device A receives a status response from device B with State = 00, device A will send a REQ request containing data to device B. If device B does not respond within 50ms, device A will resend the request.

[0073] When device B receives data from device A, it first verifies the data and then sends the result back to device A via an ACK response. If the data verification fails, the State bit is set to 11, and device A will repeat process 3. Otherwise, the State bit is set to 10, and one round of requests ends.

[0074] Figure 8 Figures a through d illustrate the process of constructing a network using Zigbee and optimizing the network using a neuromorphic network optimization algorithm for the IoT array receiver of the power-sensitive transient electromagnetic instrument.

[0075] One possible spatial arrangement of the IoT array receiver for a power transient electromagnetic instrument is as follows: Figure 8 As shown in -a. Once all receivers are deployed and powered on, each receiver will simulate the growth of neuronal synapses, sending connection requests and receiving responses via Zigbee. After one round of request-response, the two receivers that completed the request will record each other's MAC address and signal strength. Through several rounds of loops and information exchange, the receivers will build a local area network dictionary. This dictionary records the MAC codes, connectivity, signal strength, and connection counts of other receivers in the area, and uses this dictionary to initially construct a network such as... Figure 8 The unordered network shown in the diagram (-b) represents two communicable receiver links, with the numbers indicating the number of links a receiver can establish. The receiver then optimizes its own links based on the number of links from surrounding receivers and constructs a routing table. This routing table records the message transmission paths of the receivers; these paths are fixed links and do not participate in the connection optimization process. A receiver is considered faulty and its link record in the routing table is only removed if it fails to respond for more than 5 minutes. The network optimization process is as follows... Figure 8As shown in diagrams b, c, and d, edge receivers begin optimization first. Edge receivers prioritize preserving links with receivers that have more links, severing paths with other receivers, and then propagating these changes to other receivers to update the local network dictionary. After several rounds of preservation and severance, an edge receiver will have only one link remaining. This link will be preserved in the routing table and propagated to other receivers, and these links will no longer participate in the optimization process. Central receivers will optimize in the same way until all links in the network are in the routing table.

[0076] Figure 8 -d represents the final result of network optimization. In the diagram, black represents terminals, characterized by having only one link; gray represents nodes, characterized by simultaneously connecting multiple terminals to another node or coordinator; and white represents the coordinator, characterized by connecting several terminals and nodes. It's worth noting that the hardware and software architecture of terminals, nodes, and coordinators are completely identical, with no master-slave relationship. Therefore, the network constructed using this architecture possesses extremely high dynamic adaptability; the failure of any single receiver will not cause network paralysis.

[0077] Figure 8 e,f illustrates the process of a receiver joining an existing network.

[0078] Receivers A and B broadcast connection requests to other receivers in the network. Since receiver A is closer to a node in the diagram, it will connect to that node. Receiver B has no nodes around it, but it is close to one of the terminals. In this case, receiver B will send a connection request to that terminal, and then that terminal becomes a node, connecting receiver B to the network.

[0079] The transient electromagnetic instrument generates electromagnetic wave signals. The receiver of this invention receives the signals in real time, amplifies and filters them, then performs analog-to-digital conversion, and stores them in the memory of the digital control circuit for further digital filtering, error correction, and format conversion. Receivers within the receiving area first obtain their current geographical location and clock through the IoT module, then quickly and automatically assign numbers and roles through the Zigbee chip, establish a network, synchronize the clock information of receiver members, and execute this process periodically to ensure clock synchronization, network integrity, and the synchronization of transient electromagnetic data.

[0080] This invention can be connected to a transient electromagnetic instrument via an aviation socket. When the transient electromagnetic instrument sends a data read request to the receiver, the receiver transmits the processed digital signal to the transient electromagnetic instrument through the UART / GPIO interface.

[0081] Finally, each receiver packages the processed data along with location and clock parameters and sends it to an online database via the IoT module.

Claims

1. An Internet of Things (IoT) array receiver for a power-dependent transient electromagnetic instrument, characterized in that: The system includes a main unit and a receiving coil. The receiving coil includes a coil receiving antenna (1-1) printed on a circuit board and a connector (1-3) connected by a connecting wire (1-2). The main unit includes a quick-connect socket (3), a battery pack (4), a multi-track Buck-Boost circuit (5), an analog signal processing circuit (6), a digital control circuit (7), a switch (8), and an aviation socket (9) encapsulated in a housing (2). The connector (1-3) and the quick-connect socket (3) are connected accordingly. The quick-connect socket (3) is connected to the analog signal processing circuit (6). The analog signal processing circuit (6) transmits analog signals to the analog signal processing circuit (6). The output terminal of the analog signal processing circuit (6) is connected to the digital control circuit (7), the filter module in the analog signal processing circuit (6) is connected to the coaxial interface in the aviation socket (9), and the external port of the digital control circuit (7) is connected to the aviation socket (9); the power supply circuit includes the battery pack (4) and the multi-rail Buck-Boost circuit (5), the power supply terminals of the analog signal processing circuit (6) and the digital control circuit (7) are connected to the power output terminal of the multi-rail Buck-Boost circuit (5), and the charging management chip in the multi-rail Buck-Boost circuit (5) is connected to the switch (8) and the charging interface in the aviation socket (9); The digital control circuit (7) includes: a Cortex-A35 processor and an FPGA coprocessor connected to each other. The FPGA coprocessor is connected to the digital signal output terminal of the analog signal processing circuit (6), the IoT module, the Zigbee chip, and the attitude sensor. The Cortex-A35 processor is connected to the memory chip, the RAM chip, the RG45 network signal transformer, and the Ethernet PHY chip. The Cortex-A35 processor is used to run an operating system to realize module communication, resource scheduling, and data processing. The FPGA coprocessor is used to bridge the Cortex-A35 processor, the IoT module, the Zigbee chip, and the attitude sensor. Sensors and analog signal processing circuits (6); the IoT module is used for accessing the wireless Internet, positioning, and clock calibration; the Zigbee chip is controlled by the FPGA coprocessor to realize the construction of a local self-organizing network, forming a receiver array, and cooperating with the IoT module for precise positioning and data exchange; the attitude sensor is used to assist positioning and determine the exploration direction when receiving signals; the memory chip is used for storing system programs; the memory chip is an external extended memory space for the Cortex-A35; the RG45 network signal transformer and Ethernet PHY chip are used to bridge the RG45 interface network and the Cortex-A35 to realize wired network connection; The analog signal processing circuit (6) includes: a precision instrument amplifier, a filter circuit and a high-speed analog-to-digital converter connected in sequence, wherein the precision instrument amplifier is connected to a quick-connect socket (3).

2. The IoT array receiver for the power-sensitive transient electromagnetic instrument according to claim 1, characterized in that: The aviation socket includes: a coaxial interface, a differential interface, a USB interface, a UART interface, an RG45 interface, a function enable interface, and a charging interface. The coaxial interface serves as the output terminal for transient electromagnetic analog signals, and the differential interface serves as the input terminal for system expansion synchronization signals. The USB interface is used to connect other receiver hosts and computers. The UART interface is used to transmit processed transient electromagnetic digital signals. The RG45 interface is used to connect the receiver to a wired network. The function enable interface is used to externally configure the receiver's function mode. The charging interface is used to power and charge the receiver.

3. The IoT array receiver for the power-sensitive transient electromagnetic instrument according to claim 1, characterized in that: The host adopts a modular circuit board stacking design. The circuit boards containing the battery pack (4), the multi-track Buck-Boost circuit (5), the analog signal processing circuit (6), and the digital control circuit (7) are stacked on top of each other. The quick-connect socket (3), the switch (8), and the aviation socket (9) are set on the outer shell (2).

4. The IoT array receiver for the power-sensitive transient electromagnetic instrument according to claim 1, characterized in that: Multiple receivers are placed in different locations and then synchronously receive signals from a single source from different angles.

5. The IoT array receiver for the power-sensitive transient electromagnetic instrument according to claim 1, characterized in that: The filtering circuit employs an adaptive mean and Gaussian mixture filtering algorithm for transient electromagnetic signal sequences. Perform filtering. This represents the nth sample value from the mth superposition sampling. This indicates instrument intermodulation interference noise. Indicates instrument thermionic stray noise; the steps are as follows: (1) Sampling: Signal sequence of target line l Divide the data into segments of equal time intervals, and then extract a segment of data to be processed. ,because The sequence length is very short, so it can be regarded as an additive function of a linear function and white noise; (2) Noise prediction: calculation The first moment E, the second moment D, the slope K of the linear function fitting, the energy spectral density P, and the power spectral density S are used as the basis for prediction by adding these basic parameters to a random number. Then, the prediction is used to... Baseline correction is performed to predict the first moment. of The sequence, where N represents the sequence length, is used to represent... Noise probability characteristics in the sequence; Then calculate and The joint distribution and independence are used to verify the reliability of the prediction. If it is unreliable, new random numbers are generated and added to the previous basic parameters, and the prediction is repeated until it is reliable. Each row of the sequence is sampled and noise is predicted, ultimately yielding... and ; (3) Estimating the likelihood value: Based on the sum of the sampling sequences in each superimposed sequence and energy ratio First moment, Calculate the value of the second moment of the convolution kernel The likelihood values ​​of the three parameters a, b, and c are recalculated based on the obtained values ​​of a, b, and c each time. ; (4) Find the convolution: For Calculate the average and get Then with convolution kernel Perform convolution to obtain the filtered result; (5) Repeat steps (1) to (4) to complete. Noise processing is performed on each line to complete the filtering.

6. The IoT array receiver for the power-sensitive transient electromagnetic instrument according to claim 1, characterized in that: The Zigbee chip uses the Zigbee simple chain protocol. The protocol's data packet structure includes nine data areas: MAC-R, MAC-T, REQ / ACK, State, Frame, DS, Data, DE, and CRC. These represent the target device's MAC address, the current device's MAC address, the response status, the device status, the data frame, the start of data flag, the data content, the end of data flag, and the checksum, respectively. When the REQ / ACK data bit is 0, it represents REQ: requesting the target device to communicate; when it is 1, it represents ACK: responding to the target device's request. State represents several device states: (00, Ready) can receive data, (01, Busy) busy and unable to receive data, (10, Finish) data received successfully, and (11, Retrans) data received successfully but incompletely, requesting retransmission.

7. The IoT array receiver for the power-sensitive transient electromagnetic instrument according to claim 6, characterized in that: The communication process of the Zigbee simple chain protocol includes the following steps: (1) Device A sends a REQ request to Device B with State=00. If no response is received from Device B after the set time, Device A will resend the request. (2) When device B receives the request, it will immediately send an ACK response to device A according to its own status. If device B is busy, it will return State=01 and will not respond to device A's request for a set period of time. Otherwise, it will return State=00 and then establish communication with device A. (3) When device A receives a status response from device B and State=00, device A will send a REQ request containing data to device B; if device B does not respond within the set time, device A will resend; (4) When device B receives data from device A, it first verifies the data and then sends the result to device A via ACK response. If the data verification fails, the state bit State=11, and device A will repeat step (3). Otherwise, State=10, and one round of requests ends.

8. The IoT array receiver for the power-sensitive transient electromagnetic instrument according to claim 1, characterized in that: The process of constructing a network using Zigbee among multiple receivers and optimizing the network using a neuromorphic network optimization algorithm includes: (1) Each receiver simulates the growth of neuronal synapses, sends connection requests to the outside and receives responses through the Zigbee chip, and records the other party's MAC address and signal strength; (2) Through several rounds of cycles and information exchange, the receiver establishes a local area network dictionary and initially constructs an unordered network through the dictionary; the dictionary records the MAC code, connectivity, signal strength, and number of connections of other receivers in the area; the lines represent receiver links that can communicate, and the numbers represent the number of links that the receiver can establish. (3) The receiver optimizes its own links based on the number of links of the surrounding receivers and builds a routing table to record information and pass the links; the receivers at the edge will start the optimization first, prioritizing the retention of links with receivers with more links, removing paths with other receivers, and then passing these changes to other receivers to update the local network dictionary; after several rounds of retention and removal processes, the edge receiver will have only one link left, which will be retained in the routing table and passed to other receivers, and these links will no longer participate in the optimization process; the receivers at the center will optimize in the same way until all links in the network are in the routing table; (4) When a new receiver joins an existing network, it will broadcast a connection request to the receivers in the network and assign it to a node or coordinator; if the receiver is close to an existing node, it will connect to that node; if there are no nodes around but it is close to one of the terminals, it will send a connection request to that terminal, and then that terminal will become a node and connect the receiver to the network.

Images