A high-precision electronic capsule networking hybrid positioning method

Abstract

The present invention provides a high-precision electronic capsule network hybrid positioning method. The electronic capsule includes a Bluetooth transceiver module and an ultrasonic transceiver module. The method comprises the following steps: Step 1: Place multiple electronic capsules at an initial position outside the human body; activate the Bluetooth transceiver module to form a mesh network; Step 2: Measure the distance between each electronic capsule, and each electronic capsule uploads the distance information to a host computer; The host computer calculates the absolute position information of each electronic capsule; Step 3: After the electronic capsules sequentially enter the human body, activate the ultrasonic transceiver module and construct an electromagnetic wave propagation model inside the human body to simulate electromagnetic wave transmission; Step 4: Fusion and correction of the Bluetooth positioning information and ultrasonic positioning information are performed to obtain the electronic capsule position information. The present invention utilizes multi-capsule network collaborative positioning, eliminating the reliance on external base stations for capsule positioning. A hybrid positioning method is employed, with Bluetooth positioning as the primary method and ultrasonic positioning as the supplementary method, to improve positioning accuracy.

Description

Technical Field

[0001] The present invention relates to an electronic capsule positioning technology, and in particular to a high-precision electronic capsule networking hybrid positioning method. Background Art

[0002] Electronic capsules typically remain in the human body for 6-12 hours. They can be divided into different functions, such as image transmission, drug delivery, and temperature measurement, depending on the specific medical needs at different stages of diagnosis and treatment. The capsule's real-time and precise location information is one of the most important auxiliary information for detection and treatment.

[0003] Existing electronic capsule positioning technologies include radio frequency positioning (RF positioning also includes methods such as Direction of Arrival (DOA), Time of Observation (TOA), TDOA, and RSSI), magnetic positioning, and video positioning. RF positioning and magnetic positioning are currently key research areas. RSSI positioning relies on received signal strength to measure distance for positioning. This is dependent on the transmitted signal strength and the polarization of the transmitting and receiving antennas. Furthermore, electromagnetic wave loss in the complex human environment is difficult to simulate, resulting in low measurement accuracy. Magnetic positioning generally relies on magnetic field changes within the capsule's internal magnet to determine its location. This requires multiple magnetic sensors for positioning, resulting in a complex positioning system with poor interference resistance, and many other issues remain. Summary of the Invention

[0004] In view of the above-mentioned deficiencies in the prior art, the present invention provides a high-precision electronic capsule networking hybrid positioning method, which improves positioning accuracy based on the positioning and networking of multiple electronic capsules.

[0005] To achieve the above object, the present invention adopts the following technical solutions:

[0006] A high-precision electronic capsule networking hybrid positioning method, wherein the electronic capsule includes a Bluetooth transceiver module and an ultrasonic transceiver module; the method comprises the following steps:

[0007] Step 1: Place multiple electronic capsules at the initial position outside the human body; turn on the Bluetooth transceiver module and form a MESH network;

[0008] Step 2: Measure the distance between each electronic capsule. Each electronic capsule uploads the distance information to the host computer. The host computer calculates the absolute position information of each electronic capsule.

[0009] Step 3: After the electronic capsules enter the human body, they turn on the ultrasonic transceiver module and build an electromagnetic wave propagation model inside the human body to simulate electromagnetic wave transmission;

[0010] Step 4: The Bluetooth positioning information and the ultrasonic positioning information are integrated and corrected to obtain the electronic capsule location information.

[0011] Preferably, the electronic capsules communicate with each other via Bluetooth and form a MESH network. The distance between two capsules is calculated as d=c(T r -T t ), where d represents the distance from the transmitter to the receiver; c represents the propagation speed of electromagnetic waves in free space 299792458m / s; T r Indicates the time when the receiving end receives the Bluetooth signal, T t Indicates the time when the transmitter transmits the Bluetooth signal. It is inserted into the Bluetooth signal as a timestamp so that the receiver can know the transmission time.

[0012] Preferably, the host computer uses the MDS algorithm to convert the distance between the electronic capsules into relative position information, and then calculates the absolute position information of the electronic capsules based on the relative position information. The specific steps include: assuming that there are N electronic capsules, the three-dimensional coordinates of the i-th electronic capsule are [X i ,Y i ,Z i ], then the Euclidean distance between the i-th electronic capsule and the j-th electronic capsule is d i,j

[0013]

[0014] Then the related Euclidean matrix is D

[0015]

[0016] Then define the binary centralization operation

[0017]

[0018] D ⊙m represents the mth Hadamard power of D, J is the centralization matrix and is expressed as: Among them, I N represents the N×N identity matrix, 1 Nj Represents a column vector with N elements all set to 1; Dimensionality reduction technology is used to solve the above conditions: First, The eigenvalue decomposition is performed as follows where Λ=diag(λ1,λ2,…,λ N ),λ1≥λ2≥…≥λ N for The diagonal matrix composed of the eigenvalues of N ]for The eigenvector matrix composed of the eigenvectors corresponding to the eigenvalues of The three largest eigenvalues λ1, λ2, and λ3 form a diagonal matrix Λ3 = diag(λ1, λ2, λ3), and the corresponding eigenvectors form an eigenvector matrix Q3 = [q1, q2, q3]; the absolute coordinate matrix Y of the capsule is expressed as:

[0019] Preferably, the magnetic permeability in the human body is regarded as the magnetic permeability in a vacuum. Based on this parameter and the electromagnetic wave propagation formula, the electromagnetic wave propagation distance in the human body is: where d i represents the distance that electromagnetic waves pass through tissue; v i Indicates the propagation speed of electromagnetic waves in tissues; T i represents the time it takes for an electromagnetic wave to pass through the tissue, c represents the propagation speed of the electromagnetic wave in free space, and ε ri Represents the dielectric constant of different human tissues; then build a model based on the distribution of human tissues; use this model as the boundary condition of electromagnetic wave transmission to simulate the transmission of electromagnetic waves, according to the formula Constraints Where n represents the number of tissues passed through each human tissue, and the actual distance d between the two capsules is calculated. 总 , thereby improving the positioning accuracy of the electronic capsule.

[0020] Preferably, the ultrasonic transceiver module is turned on after the capsule enters the human body, and the electronic capsules send ultrasonic signals in sequence. By calculating the time difference between different electronic capsules receiving ultrasonic signals, the time information is uploaded to the host computer via Bluetooth, and the host computer performs unified calculations to obtain the relative positions of each capsule based on ultrasonic positioning.

[0021] Preferably, the Bluetooth networking positioning information and the ultrasonic positioning information are fused, and before data fusion, a Grubbs outlier test method is used to remove singular points.

[0022] Preferably, the position of the electronic capsule is corrected using a spacing constraint adaptive weighting algorithm, and the specific method is as follows:

[0023] Let the ultrasonic positioning result obtained in one ultrasonic positioning measurement time be:

[0024] L 超声波 =(X,Y,Z),

[0025] During this period

[0026] L 蓝牙 (X,Y,Z)={(X1,Y1,Z1),(X2,Y2,Z2),...,(X n ,Y n ,Z n )},

[0027] Find the average weighted coordinates of Bluetooth positioning

[0028]

[0029] At this time, a distance threshold δS is added to determine the ultrasonic positioning coordinates L 超声波 Weighted coordinates with Bluetooth The distance d is compared with the distance threshold δS, and the positioning weights of the two are adaptively determined; the adaptive weighting rule is:

[0030]

[0031] L 融合 This is the final positioning coordinate after fusion.

[0032] Beneficial effects: The present invention mainly uses Bluetooth positioning, supplemented by ultrasonic positioning, to achieve sub-centimeter-level real-time positioning accuracy. Due to the large differences in dielectric constants of different tissues in the human body, the positioning error using Bluetooth is large, at around the centimeter level. The wavelength of ultrasound in the human body is about 3.6-3.97 cm (40kHz). By using ultrasound to measure the distance between each capsule, the shortcomings of Bluetooth positioning can be compensated. The present invention constructs an electromagnetic wave propagation model inside the human body as a boundary condition, and then improves the positioning accuracy of the capsule through a network fitting algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 This is a schematic diagram of capsule networking;

[0034] Figure 2 is the reference point of the capsule’s initial position;

[0035] Figure 3 is the difference in dielectric constant of human tissue;

[0036] Figure 4 It is the overall flow chart of the present invention. DETAILED DESCRIPTION

[0037] The electronic capsule of the present invention has a built-in Bluetooth transceiver module and an ultrasonic transceiver module, and the number of electronic capsules is not less than 5. Figure 1 Specifically, multiple electronic capsules are fixed in sequence at positions with known coordinates, such as placing a capsule box at the position where the human navel is close to the skin as the initial position reference point, such as Figure 2 As shown, it is fixed at this position to match the electromagnetic wave propagation model constructed later, and this position is used as the center origin of the model to construct the model. Then the Bluetooth modules of each capsule are turned on, and the electronic capsules communicate with each other through Bluetooth and form a MESH network, and the arrival time (TOA) technology (according to the formula d=c(T r -Tt ), where d represents the distance from the transmitter to the receiver; c represents the propagation speed of electromagnetic waves in free space 299792458m / s; T r Indicates the time when the receiving end receives the Bluetooth signal, T t Indicates the time when the transmitter transmits the Bluetooth signal. The Bluetooth signal is inserted as a timestamp so that the receiver can know the transmission time. ) Calculate the distance between the two capsules. Each electronic capsule then uploads the distance information to the host computer. The host computer uses the MDS algorithm to convert the distance between the capsules into relative position information. Figure 1 As shown, assuming there are N electronic capsules, the three-dimensional coordinates of the i-th electronic capsule are [X i ,Y i ,Z i ], then the Euclidean distance between the i-th electronic capsule and the j-th electronic capsule is d i,j

[0038]

[0039] Then the related Euclidean matrix is D

[0040]

[0041] Then define the binary centralization operation

[0042]

[0043] D ⊙m represents the mth Hadamard power of D, J is the centralization matrix and can be expressed as: Among them, I N represents the N×N identity matrix, 1 Nj Represents a column vector with N elements all set to 1. Dimensionality reduction technology is used to solve the above condition optimization. First, The eigenvalue decomposition is performed as follows where Λ=diag(λ1,λ2,…,λ N ),λ1≥λ2≥…≥λ N for The diagonal matrix composed of the eigenvalues of N ]for The eigenvector matrix composed of the eigenvectors corresponding to the eigenvalues of . The three largest eigenvalues λ1, λ2, and λ3 form a diagonal matrix Λ3 = diag(λ1,λ2,λ3), and the corresponding eigenvectors form an eigenvector matrix Q3 = [q1,q2,q3]. Then the absolute coordinate matrix Y of the capsule can be expressed as:

[0044] The absolute position information of multiple electronic capsules was obtained in the previous step. When the capsules enter the human body, the dielectric constants of various tissues and organs in the human body are inconsistent and vary greatly, such as Figure 2 As shown, this will cause the propagation speed of electromagnetic waves in different parts of the human body to be different, thereby increasing the TOA ranging error. Therefore, it is necessary to build an electromagnetic wave propagation model inside the human body.

[0045] An example of the constructed electromagnetic wave propagation model is:

[0046] At the Bluetooth frequency of 2.4 GHz, the electrical parameters of some human tissues are shown in Table 1. The magnetic permeability in the human body can be regarded as the magnetic permeability in a vacuum. Based on this parameter and the electromagnetic wave propagation formula, the electromagnetic wave propagation in human tissue is: where d i represents the distance that electromagnetic waves pass through tissue; v i Indicates the propagation speed of electromagnetic waves in tissues; T i represents the time it takes for an electromagnetic wave to pass through the tissue, c represents the propagation speed of the electromagnetic wave in free space, and ε ri The dielectric constants of different human tissues are expressed. The model is then constructed based on the distribution of human tissues (human tissue map).

[0047] Human tissue or organ <![CDATA[Relative dielectric constant ε r > skin <![CDATA[ε r1 38.06]]> muscle <![CDATA[ζ r2 52.79]]> blood <![CDATA[ζ r3 58.35]]> liver <![CDATA[ζ r4 54.81]]> stomach <![CDATA[ζ r5 62.24]]> kidney <![CDATA[ζ r6 52.86]]> Fat <![CDATA[ζ r7 5.28]]>

[0048] This model is used as the boundary condition of electromagnetic wave transmission to simulate the transmission of electromagnetic waves. According to the formula Constraints Where n represents the number of tissues passed through each human tissue, the actual distance d between the two capsules can be calculated 总 , thereby improving the positioning accuracy of the electronic capsule.

[0049] The ultrasonic transceiver module is turned on after the capsule enters the human body, and the capsules send ultrasonic signals in turn to measure the distance between the capsules. Specifically, the TDOA positioning method can be applied (by calculating the time difference between different electronic capsules receiving ultrasonic signals, the distance between the positioning target and each receiving end is determined, such as Figure 1 As shown, taking capsule 1 as an example, it can be expressed as:

[0050]

[0051] Among them, X1, Y1, Z1 are the three-dimensional coordinates of capsule 1, X2, Y2, Z2; X3, Y3, Z3; X4, Y4, Z4 are the three-dimensional positions of the other four capsules respectively, d 1,2 ;d 1,3 ;d 1,4 is the distance between the four receiving ends and the transmitting end.

[0052] ΔT 23 ,ΔT 24 ,ΔT 34 Represents the TDOA between the four receiving ends, namely:

[0053] ΔT ij =T i -T j ; 2≤i,j≤4, where i and j represent the i-th and j-th receivers, respectively. This allows us to construct a set of equations about the target coordinates, yielding the following formula:

[0054]

[0055] When V is known, the target can be located by solving the above formula.

[0056] Each electronic capsule sends an ultrasonic signal in turn and records the time of receiving the ultrasonic signal sent by other capsules. The time information is then uploaded to the host computer via Bluetooth, and the host computer performs unified calculations to obtain the relative positions of each capsule based on ultrasonic positioning.

[0057] Finally, the Bluetooth network positioning information is fused with the ultrasonic positioning information. Before data fusion, in order to make the calculation results more accurate, it is necessary to remove the singular points. The Grubbs outlier test method is used. For example, for a certain capsule [X i ,Y i ,Z i ] Collect 20 times of location information, {(X i1 ,Y i1 ,Z i1 ),(X i2 ,Y i2 ,Z i2 ),…(X i20 ,Y i20 ,Z i20 )}, calculate its average position information and the standard deviation of the distance Recalculate Then G it Compared with the critical value GP(n) given by the Grubbs table, if the calculated If the value is greater than the critical value GP(n) in the table, it can be judged that the measured data is an abnormal value and can be eliminated. The critical value GP(n) is related to two parameters: the confidence probability P and the number of measurements n. Here, P = 0.95, n = 20, and G95(20) = 2.557. Therefore, if It can be considered as an outlier and removed.

[0058] After the above steps are completed, the position of the capsule is corrected using the spacing constraint adaptive weighting algorithm to obtain high-precision electronic capsule positioning information.

[0059] The specific method is as follows:

[0060] Since the output frequency of Bluetooth positioning results is higher than that of ultrasonic positioning results, generally speaking, Bluetooth is 10 times per second and ultrasonic is 3-5 times per second, so in actual dynamic positioning, an average weighted method can be adopted. Let the ultrasonic positioning result obtained in one ultrasonic positioning measurement time be:

[0061] L 超声波 =(X,Y,Z),

[0062] During this period

[0063] L 蓝牙 (X,Y,Z)={(X1,Y1,Z1),(X2,Y2,Z2),...,(X n ,Y n ,Z n )}, calculate the average weighted coordinates of Bluetooth positioning

[0064]

[0065] At this time, a distance threshold δS is added to determine the ultrasonic positioning coordinates L 超声波 Weighted coordinates with Bluetooth The distance d between the two is compared with the distance threshold δS, and the positioning weights of the two are adaptively determined. When d≤δS, the positioning results of the two are considered close, indicating that the positioning results of ultrasonic and Bluetooth are both within the normal range; when d>δS, the positioning results of the two are quite different, and there may be a large positioning error, or the deviation directions of the two positioning are exactly opposite. Therefore, the following adaptive weighting rule can be established

[0066]

[0067] L 融合 This is the final positioning coordinate after fusion. The distance threshold δS is determined by the environment and the positioning error between the two, and is set to 1 cm here.

Claims

1. A high-precision electronic capsule network hybrid positioning method, characterized in that: The electronic capsule includes a Bluetooth transceiver module and an ultrasonic transceiver module; the method includes the following steps: Step 1: Place multiple electronic capsules at the initial position outside the human body; turn on the Bluetooth transceiver module and form a MESH network; Step 2: Measure the distance between each pair of electronic capsules. Each electronic capsule uploads the distance information to the host computer. The host computer calculates the absolute position information of each electronic capsule. Step 3: After the electronic capsules enter the human body, they turn on the ultrasonic transceiver module and build an electromagnetic wave propagation model inside the human body to simulate electromagnetic wave transmission; Step 4: The Bluetooth positioning information and the ultrasonic positioning information are integrated and corrected to obtain the electronic capsule location information.

2. A high-precision electronic capsule network hybrid positioning method according to claim 1, characterized in that: The electronic capsules communicate with each other via Bluetooth and form a MESH network. The distance between two capsules is calculated as d = c (T r -T t ), where d represents the distance from the transmitter to the receiver; c represents the propagation speed of electromagnetic waves in free space 299792458m / s; T r Indicates the time when the receiving end receives the Bluetooth signal, T t Indicates the time when the transmitter transmits the Bluetooth signal. It is inserted into the Bluetooth signal as a timestamp so that the receiver can know the transmission time.

3. The high-precision electronic capsule network hybrid positioning method according to claim 1, characterized in that: The host computer uses the MDS algorithm to convert the distance between the electronic capsules into relative position information, and then calculates the absolute position information of the electronic capsules based on the relative position information. The specific steps include: assuming there are N electronic capsules, the three-dimensional coordinates of the i-th electronic capsule are [X i ,Y i ,Z i ], then the Euclidean distance between the i-th electronic capsule and the j-th electronic capsule is d i,j Then the related Euclidean matrix is D Then define the binary centralization operation D ⊙m represents the mth Hadamard power of D, J is the centralization matrix and is expressed as: Among them, I N represents the N×N identity matrix, 1 N Represents a column vector with N elements all set to 1; Dimensionality reduction technology is used to solve the above conditions: First, The eigenvalue decomposition is performed as follows where Λ=diag(λ1,λ2,…,λ N ),λ1≥λ2≥…≥λ N for The diagonal matrix composed of the eigenvalues of N ]for The eigenvector matrix composed of the eigenvectors corresponding to the eigenvalues of The three largest eigenvalues λ1, λ2, and λ3 form a diagonal matrix Λ3 = diag(λ1, λ2, λ3), and the corresponding eigenvectors form an eigenvector matrix Q3 = [q1, q2, q3]; the absolute coordinate matrix Y of the capsule is expressed as:

4. A high-precision electronic capsule network hybrid positioning method according to claim 2, characterized in that: The magnetic permeability in the human body is regarded as the magnetic permeability in a vacuum. Based on this parameter and the electromagnetic wave propagation formula, the electromagnetic wave propagation distance in the human body is: where d i represents the distance that electromagnetic waves pass through tissue; v i Indicates the propagation speed of electromagnetic waves in tissues; T i represents the time it takes for electromagnetic waves to pass through tissue, ε ri Represents the dielectric constant of different human tissues; then build a model based on the distribution of human tissues; use this model as the boundary condition of electromagnetic wave transmission to simulate the transmission of electromagnetic waves, according to the formula Constraints Where n represents the number of tissues passed through, and the actual distance d between the two capsules is calculated. 总 .

5. The high-precision electronic capsule network hybrid positioning method according to claim 1, characterized in that: The ultrasonic transceiver module is turned on after the capsule enters the human body. The electronic capsules send ultrasonic signals in turn. By calculating the time difference between different electronic capsules receiving ultrasonic signals, the time information is uploaded to the host computer via Bluetooth. The host computer then calculates the relative positions of each capsule based on ultrasonic positioning.

6. The high-precision electronic capsule network hybrid positioning method according to claim 1, characterized in that: The Bluetooth network positioning information and the ultrasonic positioning information are fused. Before data fusion, the Grubbs outlier test method is used to remove singular points.

7. The high-precision electronic capsule network hybrid positioning method according to claim 1, characterized in that: The position of the electronic capsule is corrected using the spacing constraint adaptive weighting algorithm. The specific method is as follows: Let the ultrasonic positioning result obtained in one ultrasonic positioning measurement time be: L 超声波 =(X,Y,Z), During this period L 蓝牙 (X,Y,Z)={(X1,Y1,Z1),(X2,Y2,Z2),...,(X n ,Y n ,Z n )}, Find the average weighted coordinates of Bluetooth positioning At this time, a distance threshold δS is added to determine the ultrasonic positioning coordinates L 超声波 Weighted coordinates with Bluetooth The distance d is compared with the distance threshold δS, and the positioning weights of the two are adaptively determined; the adaptive weighting rule is: L 融合 This is the final positioning coordinate after fusion.

Images