Intelligent lighting control method based on environmental parameters and solar street lamp

Abstract

The application discloses a kind of based on environmental parameter's wisdom lighting control method and solar street lamp, specifically related to energy-saving lighting equipment control technical field, including during night lighting period, photovoltaic panel is cut into weak light reference path from charging path, the same sampling clock is used to perform time binding to photovoltaic weak light voltage value and light head light value, and supply state bit is written by battery end voltage, output double-source sampling item;During LED basic lighting, rewrite drive duty cycle in the order of rising bit connection flat bit and flat bit connection falling bit, read near-field reflection value by triggering each duty cycle rewrite, so that duty cycle rewrite bit is generated through reflection response to accept duty cycle reflection item;During night lighting, weak light response of photovoltaic panel, light head light response and LED duty cycle disturbance response are collected at the same time, and environmental acceptance, light head disturbance and LED self-luminous response are separated by cross-correlation operation, then LED lighting control result is generated by Kalman recursion estimation.

Description

Technical Field

[0001] This invention relates to the field of energy-saving lighting equipment control technology, and more specifically, to a smart lighting control method based on environmental parameters and a solar street light. Background Technology

[0002] In the lighting control of solar street lights, existing solutions mostly aim to match nighttime lighting with the remaining capacity of the battery. The photovoltaic panels charge the battery during the day, and after nightfall, the LED light source is controlled by switching, dimming, or triggering brightness enhancement based on the photosensitive components on the lamp head, human body induction, vehicle induction, and battery voltage. In independent power supply scenarios such as rural roads, park branch roads, and the outer edge of parking lots, a single lamp needs to rely on limited energy storage to complete the lighting for the whole night. The lamps are also limited by cost, power consumption, installation space and maintenance frequency, making it difficult to install cameras, visibility meters or multi-point illuminance detection devices for a long time. Under these conditions, the photosensitive element on the lamp head side is easily affected by headlight sweeping, dust accumulation on the lamp cover, water film adhesion, wet road reflection, and near-field stray light. This can be directly observed on-site: when the road boundary has not become clear, the reading on the lamp head side has already increased briefly and caused a decrease in brightness; when fog or cloud reflection changes the environmental reading, the pedestrian crossing area is still not clearly identifiable; when local external light enters the vicinity of the lamp head, the LED output is incorrectly suppressed, and then it is subsequently brightened again due to pedestrians and vehicles, resulting in ineffective energy storage consumption. In the above process, the weak light response of the photovoltaic panel facing the open sky at night was not used to judge the lighting, and the structural characteristics of solar street lights compared with ordinary mains lights were not used to correct the distortion of environmental parameters on the lamp head side. The technical problem to be solved by this application is: how to use the nighttime low light response of the existing photovoltaic panels of solar streetlights to correct the environmental parameters on the lamp head side without adding high-power external detection equipment, and generate reliable LED lighting control results accordingly. Summary of the Invention

[0003] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a smart lighting control method and a solar street light based on environmental parameters. By simultaneously collecting the weak light response of the photovoltaic panel, the light-receiving response of the lamp head, and the duty cycle disturbance response of the LED during nighttime lighting, and separating the environmental reception, lamp head disturbance, and LED self-emission response through cross-correlation calculation, the LED lighting control result is generated through Kalman recursive estimation, thereby solving the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a smart solar street light based on environmental parameters, comprising: The dual-source acquisition module is used to switch the photovoltaic panel from the charging path to the low-light reference path during nighttime lighting. It binds the photovoltaic low-light voltage value and the light received value of the lamp head through the same sampling clock, and writes the power supply status bit from the battery terminal voltage to output dual-source sampling items. The LED duty cycle disturbance sampling module is used to rewrite the drive duty cycle during LED basic lighting in the order of rising position followed by level position and level position followed by falling position. It triggers the reading of near-field reflection value by rewriting the duty cycle each time, so that the duty cycle rewritten bit is used to generate the duty cycle reflection term through reflection response. The cross-correlation operation separation module is used to perform hysteresis shift, term-by-term multiplication, and bit-by-bit accumulation on the dual-source sampling terms through cross-correlation operation to generate the first correlation string; the first correlation string generates the sky receiving field through adjacent hysteresis bit sign continuation determination, and generates the lamp head disturbance field through sign foldback determination; the duty cycle reflection term generates the second correlation string through the same cross-correlation operation, and the second correlation string generates the LED self-illumination field through adjacent hysteresis bit sign continuation determination; The Kalman recursive estimation module is used to generate the current state prediction vector by multiplying the state correction vector of the previous time step with the state transition matrix through Kalman recursive estimation. The sky receiving field and LED self-illumination field are written into the observation vector after field ordering, and the lamp head disturbance field is written into the observation noise matrix after squaring. The state prediction vector is used to generate the current state correction vector through residual calculation, Kalman gain calculation and state correction.

[0005] In a preferred embodiment, it further includes: The energy-saving lighting equipment drive control module is used to solve the environmental state difference and LED state difference based on the current state correction vector, and combine them with the power supply status bit to form a duty cycle rewrite amount; when the environmental state difference is positive and the LED state difference is in the direction of the duty cycle disturbance, the LED duty cycle reduction amount is written; when the lamp head disturbance field corresponds to the sign foldback judgment, the LED duty cycle reduction amount is deleted; when the LED state difference deviates from the direction of the duty cycle disturbance, the light-emitting branch duty cycle reduction amount is written, and the lighting control result is output. The intelligent lighting feedback execution module includes a controller duty cycle output terminal, a constant current driver chip, a branch switch array, LED light-emitting branches, a current sampling resistor, and a readback sampling terminal. It is used to receive the lighting control results from the controller duty cycle output terminal and output the drive duty cycle level. The constant current driver chip generates the branch drive current based on the drive duty cycle level. The branch switch array connects the corresponding LED light-emitting branch. The current sampling resistor converts the branch drive current into a sampling voltage. The readback sampling terminal generates a drive readback value based on the sampling voltage and sends it to the LED duty cycle disturbance sampling module to participate in the next round of duty cycle reflection term generation, forming the intelligent lighting result of the solar street light.

[0006] In a preferred embodiment, the dual-source acquisition module includes: During nighttime lighting, the charging cut-off flag and lighting start flag are read. When both the charging cut-off flag and the lighting start flag are valid, a nighttime switching bit is generated. When the nighttime switching bit is valid, the nighttime switching circuit disconnects the charging branch between the photovoltaic panel and the charging control terminal, and connects the weak light reference branch between the photovoltaic panel and the weak light sampling terminal. The weak light sampling terminal reads the photovoltaic panel terminal voltage and completes the connection confirmation, and outputs the photovoltaic reference connection bit. The same sampling clock is started based on the photovoltaic reference access point, and the sampling clock generates a time number. Under the same time number, the photovoltaic panel voltage is written into the photovoltaic weak light voltage value through analog-to-digital conversion, and the lamp head light receiving voltage is written into the lamp head light receiving value through analog-to-digital conversion. When the photovoltaic weak light voltage value or lamp head light receiving value is missing under the same time number, the missing field is written into the empty sampling field, and the output time is bound to the sampling item. Read the battery terminal voltage at the same time number as the time-bound sampling item, subtract the battery terminal voltage at the previous time number from the battery terminal voltage to obtain the terminal voltage difference; write the positive terminal voltage difference to the power supply up bit, write the zero terminal voltage difference to the power supply level bit, and write the negative terminal voltage difference to the power supply down bit, and merge the corresponding power supply status bit into the time-bound sampling item to output the dual-source sampling item.

[0007] In a preferred embodiment, the LED duty cycle disturbance sampling module includes: During LED basic lighting, the basic duty cycle value and duty cycle step value are read. The rising duty cycle value is obtained by adding the duty cycle step value to the basic duty cycle value. The level duty cycle value is written to the basic duty cycle value. The falling duty cycle value is obtained by subtracting the duty cycle step value from the basic duty cycle value. The duty cycle rewrite bit sequence is formed in the order of rising followed by level and level followed by falling. The driver duty cycle is written bit by bit to the LED driver terminal according to the duty cycle rewrite bit sequence. Each time a duty cycle rewrite bit is written, a reflection sampling trigger bit is generated. The near-field reflection sampling terminal reads the near-field reflection voltage at the same sampling clock after the reflection sampling trigger bit is established, converts the near-field reflection voltage into a near-field reflection value, and outputs the duty cycle reflection candidate. Write the duty-overwrite bit, near-field reflection value, and sampling time number from the duty-over-reflection candidate to the same record item. Subtract the near-field reflection value corresponding to the equal-position duty-over-reflection value from the near-field reflection value to generate the reflection response difference. Then, bind the duty-overwrite bit to the reflection response difference and write it to generate the duty-over-reflection item.

[0008] In a preferred embodiment, the cross-correlation calculation separation module includes: Using the sampling time number in the dual-source sampling item as an index, the photovoltaic weak light voltage difference between adjacent time points is first written into the photovoltaic change position, and then the light reception difference of the lamp head between adjacent time points is written into the lamp head change position. For each lag position, the photovoltaic change position is moved to before the lamp head change position according to the sampling time number, and the product term of the photovoltaic change position and the lamp head change position is generated bit by bit. The product term is merged and accumulated according to the lag position to form the first correlation value, forming the first correlation string that is sequentially inherited by the lag positions. The first correlation string reads the first correlation value sequentially along the lag bits. When the sign of the next first correlation value is the same as that of the previous first correlation value and the direction of numerical displacement has not changed, the corresponding lag bit is written into the sky receiving field. When the sign of the next first correlation value is opposite to that of the previous first correlation value and the direction of numerical displacement is reversed, the corresponding lag bit is written into the lamp head disturbance field. When the same lag bit hits both the sky receiving field and the lamp head disturbance field at the same time, the field in which the sign reverses first is retained, and the replaced field is written back as the field to be reviewed. Using the lag bits defined by the field to be reviewed as avoidance bits, the duty rewrite in the duty cycle reflection term is shifted to before the near-field reflection value. The product term of the duty cycle rewrite and the near-field reflection value is generated bit by bit, and the product term is merged and accumulated according to the lag bits to form the second correlation value. When the second correlation value is consecutive in the same direction as the duty cycle rewrite along the lag bit order, the corresponding lag bit is written to the LED self-illumination field. Lag bits that are not consecutive in the same direction are written back to the lamp head disturbance field, and the cross-correlation calculation separation record is output.

[0009] In a preferred embodiment, the Kalman recursive estimation module includes: Starting from the state correction vector of the previous moment, multiply the environmental state column and the LED state column by the corresponding columns in the state transition matrix to obtain the state prediction vector of the current moment; multiply the covariance matrix of the previous moment by the state transition matrix on the left and by the transpose on the right, and then incorporate it into the process noise matrix to obtain the prediction covariance matrix of the current moment. Based on the field order of the record separated by cross-correlation operation, the sky receiving field is written into the environmental observation position of the observation vector, and the LED self-illumination field is written into the LED observation position of the observation vector; the lamp head disturbance field is first squared by the hysteresis bit, and then written into the diagonal position of the observation noise matrix by the same hysteresis bit, so that the lamp head disturbance amount is included in the observation uncertainty term of Kalman recursive estimation.

[0010] In a preferred embodiment, the Kalman recursive estimation module further includes: Multiply the observation matrix by the current state prediction vector to obtain the predicted observation vector, and subtract the predicted observation vector from the observation vector to obtain the residual vector; substitute the observation matrix, the current prediction covariance matrix, and the observation noise matrix into the gain denominator matrix, and multiply the current prediction covariance matrix by the transpose of the observation matrix and then multiply by the inverse of the gain denominator matrix to obtain the Kalman gain matrix. The state correction is obtained by multiplying the Kalman gain matrix by the residual vector, and then added to the current state prediction vector to form the current state correction vector. When the sign bit string of the current state correction vector is the same as that of the previous state correction vector and the field order difference is zero, it is written to the recursive convergence bit. Otherwise, the current state correction vector is written back as the recursive starting point of the next time step, and the Kalman recursive estimation record is output.

[0011] In a preferred embodiment, the energy-saving lighting equipment drive control module includes: Based on the current state correction vector, read the environmental column value and the LED column value. Subtract the previous environmental column value from the current environmental column value to generate the environmental state difference, and subtract the previous LED column value from the current LED column value to generate the LED state difference. When the power supply status bit is the power supply uplink bit, write the environmental state difference into the duty cycle candidate. When the power supply status bit is the power supply neutral bit, write the overlapping part of the environmental state difference and the LED state difference with the same sign into the duty cycle candidate. When the power supply status bit is the power supply downlink bit, write the positive remainder after subtracting the LED state difference from the environmental state difference into the duty cycle candidate, and output the duty cycle candidate record. Read the duty candidate quantity along the duty candidate record. When the environmental state difference is positive and the LED state difference continues along the duty disturbance direction generated by the LED duty disturbance sampling module, write the duty candidate quantity into the LED duty reduction quantity. When the symbol of the lamp head disturbance field is returned to the determination, move the written LED duty reduction quantity into the brightness reduction deletion bit and block the brightness reduction deletion bit from entering the constant current drive writing end of the energy-saving lighting device. When the LED state difference deviates from the duty cycle disturbance direction, the absolute value of the LED state difference is multiplied by the power supply status bit to generate a branch low candidate quantity, and the branch low candidate quantity is written into the light-emitting branch duty cycle low quantity; when the LED duty cycle reduction quantity has not entered the dimming deletion bit, it is written into the total duty cycle write field, and the light-emitting branch duty cycle low quantity is written into the branch duty cycle write field. The total duty cycle write field and the branch duty cycle write field are combined to form the lighting control result.

[0012] In a preferred embodiment, the smart lighting feedback execution module includes: After the lighting control results enter the controller duty cycle output, the controller duty cycle output first reads the dimming deletion bit. If the dimming deletion bit is hit, the total duty cycle register value of the previous moment is retained. If the dimming deletion bit is not hit, the total duty cycle register value of the previous moment is subtracted from the LED duty cycle reduction amount to generate the current total duty cycle register value. The controller duty cycle output then locates the branch duty cycle register bit according to the branch number in the branch duty cycle write field, and generates the current branch duty cycle register value by subtracting the low duty cycle voltage of the light-emitting branch from the original value of the branch duty cycle register bit. The output drives the duty cycle level and the branch selection level. The constant current driver chip reads the current total duty cycle register value and rewrites the number of high-level hold bits within the drive cycle according to the current total duty cycle register value, so that the drive duty cycle level is converted into a constant current control pulse; the branch switch array reads the branch selection level and controls the conduction time slot of the corresponding LED light-emitting branch according to the current branch duty cycle register value, so that the constant current control pulse enters the connected LED light-emitting branch and the output branch outputs light execution item; A current sampling resistor is connected in series with the LED light-emitting branch that is turned on. The branch driving current passes through the current sampling resistor to form a sampling voltage. The readback sampling terminal reads the sampling voltage at the end of the same driving cycle and subtracts the sampling voltage of the previous driving cycle from the sampling voltage to generate a driving readback value. The driving readback value is written into the next round of duty cycle reflection term according to the sampling time to form the smart lighting result of the solar street light.

[0013] A smart lighting control method based on environmental parameters, the method comprising: S1. During nighttime lighting, the photovoltaic panel is switched from the charging path to the low light reference path. The photovoltaic low light voltage value and the light received value of the lamp head are time-bound through the same sampling clock, and the battery terminal voltage is written into the power supply status bit to output the dual-source sampling item. S2. During LED basic lighting, the duty cycle is rewritten in the order of rising position followed by level position and level position followed by falling position. The near-field reflection value is read by each duty cycle rewriting, so that the duty cycle rewritten position is used to generate the duty cycle reflection term through the reflection response. S3. Perform hysteresis shift, term-by-term multiplication, and bit-by-bit accumulation on the dual-source sampling terms through cross-correlation operation to generate the first correlation string; the first correlation string generates the sky receiving field through adjacent hysteresis bit sign continuation determination, and generates the lamp head disturbance field through sign foldback determination; the duty cycle reflection term generates the second correlation string through the same cross-correlation operation, and the second correlation string generates the LED self-illuminating field through adjacent hysteresis bit sign continuation determination. S4. The state prediction vector at the current moment is generated by multiplying the state correction vector of the previous moment by the state transition matrix through Kalman recursion estimation. The sky receiving field and LED self-illumination field are written into the observation vector after field ordering. The lamp head disturbance field is written into the observation noise matrix after squaring. The state prediction vector is used to generate the state correction vector at the current moment by residual calculation, Kalman gain calculation and state correction. S5. Solve the environmental state difference and LED state difference based on the current state correction vector, and combine them with the power supply state bit to form the duty cycle rewrite amount; when the environmental state difference is positive and the LED state difference is in the direction of the duty cycle disturbance, write the LED duty cycle down adjustment amount; when the lamp head disturbance field corresponds to the sign return judgment, delete the LED duty cycle down adjustment amount; when the LED state difference deviates from the direction of the duty cycle disturbance, write the light-emitting branch duty cycle low amount, and output the lighting control result.

[0014] The technical effects and advantages of this invention are as follows: This solution binds the photovoltaic low-light voltage value and the light received value of the lamp head at the same time, so that the low-light response of the photovoltaic panel at night participates in environmental correction, thereby relatively suppressing the false brightness reduction caused by vehicle headlight sweeping and wet road reflection; By sampling the LED duty cycle disturbance and near-field reflection value, the LED self-emission response can be extracted separately, thereby reducing the continuous power consumption of the ineffective light-emitting branch. Based on cross-correlation calculations, the sky reception field, lamp head disturbance field, and LED self-illumination field are separated, so that the source of environmental change can be distinguished before recursion, thereby improving the basis for lighting judgment. By incorporating the lamp head disturbance field into the observation noise matrix through Kalman recursive estimation, the local disturbance does not directly affect the environmental state, thereby improving the stability of LED duty cycle rewriting. The overall lamp brightness reduction and branch voltage reduction are processed separately based on the brightness reduction deletion bit and the duty cycle low voltage of the light-emitting branch, so that the lighting control results are more in line with the power supply status and effective light output status. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the system module structure of the present invention; Figure 2 This is a flowchart outlining the method steps of the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] Refer to the instruction manual appendix Figure 1-2 The present invention provides a smart solar street light based on environmental parameters, comprising: The dual-source acquisition module is used to switch the photovoltaic panel from the charging path to the low-light reference path during nighttime lighting. It binds the photovoltaic low-light voltage value and the light received value of the lamp head through the same sampling clock, and writes the power supply status bit from the battery terminal voltage to output dual-source sampling items. This implementation revolves around a dual-source acquisition module. In the dual-source acquisition module, the positive end of the photovoltaic panel is connected to the common input terminal of the nighttime switching circuit. The first output terminal of the nighttime switching circuit is connected to the charging control terminal, and the second output terminal of the nighttime switching circuit is connected to the low-light sampling terminal. The light-receiving end of the lamp head is connected to the lamp head sampling input terminal of the controller, and the positive end of the battery is connected to the battery sampling input terminal of the controller via a voltage divider sampling branch. The controller outputs a switching control signal to the nighttime switching circuit and outputs the same sampling clock to the low-light sampling terminal, the lamp head sampling input terminal, and the battery sampling input terminal. Through this connection structure, the photovoltaic panel charges the battery via the charging control terminal during the day, provides a low-light sky reference via the low-light sampling terminal at night, provides light-receiving input via the lamp head light-receiving end, and provides power supply change input via the battery sampling input terminal. After the three types of inputs enter the same sampling sequence, they form a dual-source sampling item. This implementation process includes the following steps: Nighttime path switching is used to allow the photovoltaic panel terminal electrical signal to enter unidirectionally between the charging branch and the low-light reference branch. The charging control terminal writes a charging cutoff flag after the charging switch is turned off and the charging branch current sampling field is empty. The controller writes a lighting start flag after the LED basic lighting current is written and the drive readback value is formed. The dual-source acquisition module reads the charging cutoff flag and the lighting start flag. When both are valid, a nighttime switching bit is generated. After the nighttime switching bit enters the nighttime switching circuit, the nighttime switching circuit first disconnects the charging branch between the photovoltaic panel positive terminal and the charging control terminal, and then connects the low-light reference branch between the photovoltaic panel positive terminal and the low-light sampling terminal, so that the two branches are not simultaneously connected. The low-light sampling terminal reads the photovoltaic panel terminal voltage. When the photovoltaic field is written with the sampling code and the charging branch current sampling field is empty, the dual-source acquisition module writes the photovoltaic reference access bit. When the photovoltaic field is empty or the charging branch current sampling field still has a current sampling value, the dual-source acquisition module writes the access failure bit and stops generating this round of dual-source sampling items. Simultaneous sampling binding ensures that the photovoltaic (PV) low-light voltage value and the lamp head's light-receiving value are included in the same sampling reference. After the PV reference access position is formed, the controller starts the same sampling clock and generates time numbers according to the sampling trigger sequence. Under the same time number, the low-light sampling terminal writes the PV panel voltage through voltage divider sampling and analog-to-digital conversion into the PV low-light voltage value, and the lamp head sampling input terminal writes the lamp head light-receiving voltage through analog-to-digital conversion into the lamp head light-receiving value. The PV low-light voltage value is written into the PV field, and the lamp head light-receiving value is written into the lamp head field. If the same time number lacks a PV low-light voltage value, an empty sampling field is written to the missing PV bit. If the same time number lacks a lamp head light-receiving value, an empty sampling field is written to the missing lamp head bit. Time-binding sampling items with empty sampling fields retain the time number and do not participate in subsequent PV change bit and lamp head change bit calculations. The dual-source acquisition module outputs time-binding sampling items. The power supply status is written to incorporate changes in battery terminal voltage into the same time number. The positive terminal of the battery is input to the battery sampling input terminal of the controller via the voltage divider sampling branch. The dual-source acquisition module reads the battery terminal voltage with the same time number as the time-bound sampling item and reads the battery terminal voltage under the previous time number. The current battery terminal voltage minus the previous battery terminal voltage gives the terminal voltage difference. When the terminal voltage difference is positive, the power supply status bit is written to the power supply uplink bit and encoded as +1. When the terminal voltage difference is zero, the power supply status bit is written to the power supply level bit and encoded as 0. When the terminal voltage difference is negative, the power supply status bit is written to the power supply downlink bit and encoded as -1. When the first time number does not have the previous battery terminal voltage, the power supply status bit is written to the initial power supply bit, and the terminal voltage difference calculation is not performed. After the power supply status bit is incorporated into the time-bound sampling item, the dual-source acquisition module outputs a dual-source sampling item containing the time number, photovoltaic weak light voltage value, lamp head light value, empty sampling field, and power supply status bit. Through the above processing, the dual-source acquisition module writes the photovoltaic panel weak light reference, lamp head light input and battery power supply changes into the same time number during nighttime lighting, and avoids path conflict by disconnecting the charging branch and the weak light reference branch before connecting them. In practical applications, after nightfall, the charging control terminal of the solar street light stops charging the battery, the controller starts the LED basic lighting, and the night switching circuit connects the photovoltaic panel to the weak light sampling terminal. When the vehicle headlight only sweeps across the vicinity of the lamp head, the light value received by the lamp head changes, but the photovoltaic weak light voltage value does not change synchronously. The dual-source sampling term retains the difference between the two for the cross-correlation calculation separation module to read. When the light value received by the lamp head is missing at a certain moment, the empty sampling field prevents the generation of lamp head change bits across the missing moment, avoiding the Kalman recursive estimation module reading mismatch data.

[0018] The LED duty cycle disturbance sampling module is used to rewrite the drive duty cycle during LED basic lighting in the order of rising position followed by level position and level position followed by falling position. It triggers the reading of near-field reflection value by rewriting the duty cycle each time, so that the duty cycle rewritten bit is used to generate the duty cycle reflection term through reflection response. This embodiment revolves around an LED duty cycle perturbation sampling module, used to write controlled duty cycle perturbations to the LED driver during LED basic lighting, and to read the near-field reflection response of the lamp head after each duty cycle rewrite, so that the LED emission change and the near-field reflection change form a duty cycle reflection term at the same sampling moment; in the LED duty cycle perturbation sampling module, the controller duty cycle output terminal is connected to the LED driver terminal, the LED driver terminal is connected to the constant current driver chip, the constant current driver chip output terminal is connected to the LED emission branch, and the near-field reflection sampling terminal is arranged in the reflection area adjacent to the LED emission branch and connected to the controller reflection sampling input terminal; this connection relationship is used to ensure that the driver duty cycle rewrite first enters the LED emission branch, and then the near-field reflection sampling terminal reads the reflection change caused by the duty cycle rewrite, so that the cross-correlation calculation separation module can identify the LED self-emission field; the implementation process includes the following steps: The duty cycle rewrite bit sequence is used to generate three distinguishable driving disturbances—rising, level, and falling—without exiting the LED base lighting. The LED duty cycle disturbance sampling module reads the base duty cycle value and the duty cycle step value. The base duty cycle value is taken from the current total duty cycle register value during the LED base lighting period, and the duty cycle step value is taken from the single-step duty cycle register difference stored in the controller. The base duty cycle value plus the duty cycle step value yields the rising duty cycle value, the base duty cycle value is written to the level duty cycle value, and the base duty cycle value minus the duty cycle step value yields the rising duty cycle value. The falling-down duty cycle value is obtained; if the rising-up duty cycle value exceeds the upper boundary of the duty cycle register that can be written to the LED driver, the rising-up duty cycle value is written to the upper boundary of the duty cycle register; if the falling-down duty cycle value is lower than the basic lighting reserved duty cycle value, the falling-down duty cycle value is written to the basic lighting reserved duty cycle value, and the corresponding rewriting result is written to the duty cycle limit field; the rising-up duty cycle value, the level duty cycle value, and the falling-down duty cycle value are written to the duty cycle rewriting bit sequence in the order of rising followed by level and level followed by falling, for subsequent bit-by-bit drive writing and reading; The single-step duty cycle register difference is generated by the total number of drive cycle count bits and the disturbance resolution level in the controller configuration table. The controller divides the total number of drive cycle count bits by the disturbance resolution level and writes the integer quotient as the single-step duty cycle register difference. The remainder of the division is not written into the single-step duty cycle register difference. The duty cycle step value reads the single-step duty cycle register difference and participates in the calculation of the rising duty cycle value and the falling duty cycle value. When the rising duty cycle value exceeds the upper boundary of the duty cycle register that the LED driver terminal can write, it is written as the upper boundary of the duty cycle register. When the falling duty cycle value is lower than the basic lighting reserved duty cycle value, it is written as the basic lighting reserved duty cycle value. The duty cycle perturbation writing is used to convert the duty cycle rewrite bit sequence bit by bit into the light output change of the LED light-emitting branch, and simultaneously obtain the near-field reflection value; the LED duty cycle perturbation sampling module reads the rising duty cycle value, the level duty cycle value, and the falling duty cycle value according to the duty cycle rewrite bit sequence, and writes them bit by bit to the LED driver terminal through the controller's duty cycle output terminal; after each duty cycle rewrite bit is written, the controller generates a reflection sampling trigger bit, and the constant current driver chip changes the number of bits of the high-level hold period of the LED light-emitting branch according to the current duty cycle rewrite bit, and the near-field reflection sampling terminal... Within the same sampling clock, the near-field reflection voltage formed by the adjacent reflection area of ​​the LED light-emitting branch is read, and the near-field reflection value is generated through analog-to-digital conversion. If the current duty cycle overwrite bit has been written but the near-field reflection value is missing, the duty cycle reflection candidate is written to the reflection missing field. If the LED driver does not read back the current duty cycle overwrite bit, the duty cycle reflection candidate is written to the driver not written field, and the generation of the reflection response difference corresponding to the duty cycle overwrite bit is stopped. The duty cycle overwrite bit, near-field reflection value, sampling time number, and exception field are all written to the duty cycle reflection candidate. The reflection response difference is used to convert the near-field reflection value from a single sample value into a response quantity that varies with the duty cycle disturbance. The LED duty cycle disturbance sampling module reads the duty cycle reflection candidate according to the sampling time number, locates the near-field reflection value corresponding to the level duty cycle value in the same group of ascending, level, and descending positions, and uses it as the reflection reference value for this group. If the level near-field reflection value is missing, the duty cycle reflection candidate in this group is marked as "no reference" and no duty cycle reflection item is generated. If there are multiple level near-field reflection values ​​in the same group, the first level near-field reflection value is retained in ascending order of the sampling time number as the reflection reference value for this group. The reflection response difference is obtained by subtracting the reflection reference value of this group from the near-field reflection value corresponding to each duty cycle rewrite bit. The duty cycle rewrite bit and the reflection response difference are bound and written according to the same sampling time number to generate a duty cycle reflection item containing the duty cycle rewrite bit, near-field reflection value, reflection response difference, sampling time number, and drive readback field, which is used by the cross-correlation calculation separation module to form a second correlation string. Through the above processing, the LED duty cycle disturbance sampling module connects the duty cycle rewriting of the LED driver, the light output change of the LED light-emitting branch, and the reflection reading of the near-field reflection sampling end into a duty cycle reflection term within the same sampling time sequence; the duty cycle limiting field prevents the rising and falling duty cycles from exceeding the driving boundary, the absence of a reference mark prevents the generation of incorrect reflection response differences when there is a lack of a level reflection reference, and the driver not written field prevents the duty cycle rewriting bits that have not actually entered the LED driver from participating in the cross-correlation calculation; In practical applications: When the solar street light is in basic lighting mode, the controller first uses the current total duty cycle register value as the base duty cycle value, and then writes the rising duty cycle value, the level duty cycle value, and the falling duty cycle value in sequence. If the light output change of the LED light-emitting branch actually affects the reflective area near the lamp head, the near-field reflection value will change accordingly with the duty cycle rewrite bit. After the duty cycle reflection term enters the cross-correlation calculation separation module, it can generate the LED self-emission field. If the level near-field reflection value is missing, the LED duty cycle disturbance sampling module will not generate this group of duty cycle reflection terms, avoiding the subsequent Kalman recursive estimation module reading the LED self-emission input without a reference.

[0019] The cross-correlation operation separation module is used to perform hysteresis shift, term-by-term multiplication, and bit-by-bit accumulation on the dual-source sampling terms through cross-correlation operation to generate the first correlation string; the first correlation string generates the sky receiving field through adjacent hysteresis bit sign continuation determination, and generates the lamp head disturbance field through sign foldback determination; the duty cycle reflection term generates the second correlation string through the same cross-correlation operation, and the second correlation string generates the LED self-illumination field through adjacent hysteresis bit sign continuation determination; This implementation revolves around a cross-correlation calculation separation module, used to sequentially separate the photovoltaic weak light changes and lamp head light reception changes in the dual-source sampling terms, and further uses the duty cycle reflection term to determine whether the near-field reflection caused by LED light emission actually exists. The cross-correlation calculation separation module is executed by the calculation unit within the controller. The calculation unit reads the dual-source sampling terms and the duty cycle reflection term from the storage area without changing the hardware connection relationship between the photovoltaic panel, the lamp head light receiving end, and the LED driver end. Its function is to first form a first correlation string from the photovoltaic change bits and the lamp head change bits, then split the first correlation string into the sky receiving field and the lamp head disturbance field, and finally form a second correlation string from the duty cycle overwrite bits and the near-field reflection value, used to generate the LED self-emission field, and continue to write back the hysteresis bits that cannot be explained by LED light emission to the lamp head disturbance field. This implementation process includes the following steps: The formation of the photovoltaic change bit, lamp head change bit, and first correlation string is used to convert the dual-source sampling items into time-series data that can be cross-correlationly analyzed. The cross-correlation analysis separation module uses the sampling time number in the dual-source sampling item as an index. First, it reads the photovoltaic weak light voltage value under adjacent valid time numbers, subtracts the photovoltaic weak light voltage value of the previous valid time number from the photovoltaic weak light voltage value of the later valid time number, and obtains the photovoltaic weak light voltage difference, which is written to the photovoltaic change bit. Then, it reads the lamp head light-receiving value under adjacent valid time numbers, subtracts the lamp head light-receiving value of the previous valid time number from the lamp head light-receiving value of the later valid time number, and obtains the lamp head light-receiving difference, which is written to the lamp head change bit. An empty sampling field is also written. The dual-source sampling items retain the sampling time number but do not participate in the adjacent difference calculation. The dual-source sampling items on both sides of the empty sampling field do not directly constitute adjacent valid time numbers. The lag bit is generated according to the number of valid time numbers entering the current round of cross-correlation operation. The shifted items that exceed the range of valid time numbers do not generate product terms. For each lag bit, the cross-correlation operation separation module shifts the photovoltaic change bit to before the lamp head change bit according to the sampling time number, so that the shifted photovoltaic change bit is multiplied bit by bit with the lamp head change bit on the same calculation bit to obtain the product term. The product terms under the same lag bit are merged and accumulated into the first correlation value. All first correlation values ​​are written in ascending order of lag bit to form the first correlation string. The formation of the sky acceptance field and the lamp head disturbance field is used to separate the same-direction acceptance change and the local disturbance change in the first correlation string; the cross-correlation operation separation module reads the first correlation value in the first correlation string in ascending order along the lag bit, and subtracts the previous first correlation value from the next first correlation value to obtain the correlation displacement value; when the correlation displacement value is positive, it is written in the upward direction; when the correlation displacement value is zero, it is written in the horizontal direction; when the correlation displacement value is negative, it is written in the downward direction; when the next first correlation value has the same sign as the previous first first correlation value, and the next correlation displacement direction is the same as the previous correlation displacement direction (both upward, both horizontal, or both downward), the cross-correlation operation separation module writes the lag bit corresponding to the next first correlation value into the sky acceptance field; When the sign of the subsequent first correlation value is opposite to that of the previous first correlation value, and the direction of the correlation displacement switches between the upward and downward directions, the cross-correlation calculation separation module writes the lag bit corresponding to the subsequent first correlation value into the lamp head disturbance field. When the same lag bit is used as both the last bit of the sky receiving field and the first bit of the lamp head disturbance field, the cross-correlation calculation separation module reads the order in which the signs reverse, retains the field in which the signs reverse first, and writes the field to be replaced along with the corresponding lag bit into the field to be verified. The field to be verified does not participate in the writing of the observation vector by the Kalman recursive estimation module until the subsequent second correlation string is verified. The formation of the second correlation string and the LED self-emissive field is used to determine whether the lag bit to be verified is caused by LED duty cycle disturbance; the cross-correlation operation separation module reads the lag bit defined by the field to be verified and uses the lag bit as the avoidance bit between the first correlation string and the second correlation string; when generating the second correlation value, the avoidance bit does not participate in the merging of the product term of the duty cycle overwrite bit and the near-field reflection value; the cross-correlation operation separation module reads the duty cycle overwrite bit and the near-field reflection value in the duty cycle reflection term, shifts the duty cycle overwrite bit to before the near-field reflection value according to the lag bit, multiplies the shifted duty cycle overwrite bit with the near-field reflection value on the same calculation bit bit bit by bit to obtain the duty cycle reflection product term, and merges and accumulates the duty cycle reflection product terms under the same lag bit to form the second correlation value. All second correlation values ​​are written in ascending order of lag bit to form the second correlation string; When the second correlation value is read in ascending order along the lag bits, if the sign continuation direction of the second correlation value forms a continuation in the same direction as the ascending, level, and descending order in the duty cycle overwrite bit sequence, the cross-correlation operation separation module writes the corresponding lag bit into the LED self-illumination field; if the second correlation value does not form a continuation in the same direction, the cross-correlation operation separation module writes the corresponding lag bit back to the lamp head perturbation field; the cross-correlation operation separation record includes at least the sky acceptance field, lamp head perturbation field, LED self-illumination field, field to be reviewed, and avoidance bit, which are used for reading by the Kalman recursive estimation module; Through the above processing, the cross-correlation calculation separation module first uses the lag multiply-add relationship between the photovoltaic weak light voltage change and the lamp head light reception change to separate the weak sky light reception and the lamp head local disturbance. Then, it uses the lag multiply-add relationship between the LED duty perturbation and the near-field reflection change to confirm the LED self-luminous response, so that the sources of change on the photovoltaic side, the lamp head side and the LED side are separated before entering the Kalman recursive estimation module. In practical applications: When cloud movement causes changes in the dim sky light, the photovoltaic change position and the lamp head change position will form a first correlation value with the same sign on adjacent lag positions. The cross-correlation operation separation module writes this lag position into the sky reception field. When the headlights only sweep across the vicinity of the lamp head, the lamp head change position changes but the photovoltaic change position does not form the same reception, and the first correlation string shows a sign reversal. The cross-correlation operation separation module writes this lag position into the lamp head perturbation field. When the near-field reflection value changes synchronously with the duty cycle rewrite position after the LED duty cycle is rewritten, the second correlation string forms a continuation result in the same direction as the duty cycle rewrite order. The cross-correlation operation separation module writes the corresponding lag position into the LED self-emission field, so that the subsequent Kalman recursive estimation module can write the LED self-emission field into the observation vector.

[0020] The Kalman recursive estimation module is used to multiply the state transition matrix by the state correction vector of the previous time step through Kalman recursive estimation to generate the state prediction vector of the current time step. The sky receiving field and LED self-illumination field are written into the observation vector after field ordering. The lamp head disturbance field is written into the observation noise matrix after squaring. The state prediction vector is used to generate the state correction vector of the current time step through residual calculation, Kalman gain calculation and state correction. This implementation revolves around a Kalman recursive estimation module. This module is located within the controller's computation unit, which connects a cross-correlation calculation separation storage area and a state recursion storage area. The resulting current-moment state correction vector is written into the read area of ​​the energy-saving lighting equipment drive control module. The state correction vector uses fixed column positions, with the environmental state column preceding the LED state column. The state transition matrix, covariance matrix, process noise matrix, and observation matrix are all generated in the row and column order according to the aforementioned column positions to avoid transposition of environmental and LED state quantities during the recursion process. This implementation includes the following steps: State prediction is used to advance the corrected environmental and LED states from the previous time step to the current time step. The Kalman recursive estimation module reads the state correction vector from the previous time step from the state recursive storage area, and reads the column values ​​according to the fixed order of the environmental state column and the LED state column. The state transition matrix is ​​given by the controller configuration table. The row and column order of the state transition matrix is ​​consistent with the column position of the state correction vector. The environmental state column and the LED state column are multiplied by the corresponding column position in the state transition matrix and accumulated row by row to obtain the state prediction vector at the current time step. The covariance matrix of the previous time step is read according to the same column position. First, the state transition matrix is ​​multiplied by the previous time step covariance matrix on the left, and then the resulting matrix is ​​multiplied by the transpose of the state transition matrix on the right. Then, it is incorporated into the process noise matrix to form the prediction covariance matrix at the current time step. The process noise matrix is ​​generated by the noise record corresponding to the state column position in the controller configuration table. When the noise record is missing, the corresponding matrix position is written with a zero value and the process noise default flag is written. The observation construct is used to convert the cross-correlation operation separation record into the observation input for Kalman recursive estimation, and to include the lamp head disturbance in the observation noise matrix. The Kalman recursive estimation module reads the cross-correlation operation separation record, writes the sky receiving field into the environmental observation position of the observation vector according to the field order, and writes the LED self-illumination field into the LED observation position of the observation vector. The field order of the observation vector is consistent with the row order of the observation matrix. The observation matrix is ​​given by the controller configuration table and corresponds to the column position of the state correction vector. The lamp head disturbance field first reads the disturbance value according to the hysteresis position, and then performs a square operation on the disturbance value. When multiple hysteresis positions belong to the same observation field, the squared values ​​are merged and accumulated according to the observation field. The accumulated result is written into the diagonal position of the corresponding observation field in the observation noise matrix, and zero value is written into the off-diagonal position. When there is a field to be verified in the cross-correlation operation separation record, the field to be verified is not written into the observation vector. The hysteresis position corresponding to the field to be verified is only entered into the observation noise matrix for merging after the LED self-illumination field verification is completed. Gain calculation is used to convert the difference between the current observation and the state prediction into the matrix required for state correction. The Kalman recursive estimation module multiplies the observation matrix by the current state prediction vector to obtain the predicted observation vector, and then subtracts the predicted observation vector from the observation vector to obtain the residual vector. Subsequently, the observation matrix is ​​multiplied on the left by the current prediction covariance matrix, then multiplied on the right by the transpose of the observation matrix, and added to the observation noise matrix to obtain the gain denominator matrix. When the gain denominator matrix has an inverse matrix, the current prediction covariance matrix is ​​multiplied on the right by the transpose of the observation matrix, and then multiplied by the inverse matrix of the gain denominator matrix to obtain the Kalman gain matrix. When the gain denominator matrix does not have an inverse matrix, the Kalman recursive estimation module writes a gain non-invertible flag, retains the current state prediction vector as the current state correction vector, and does not write the recursive convergence bit. State correction is used to form the current state correction vector for the energy-saving lighting equipment drive control module to read, and to determine the recursion starting point for the next moment. The Kalman recursion estimation module obtains the state correction amount by multiplying the Kalman gain matrix by the residual vector, and then adds the state correction amount to the current state prediction vector to form the current state correction vector. The current state correction vector is written into the state recursion storage area according to the environmental state column and the LED state column, and is simultaneously sent to the reading area of ​​the energy-saving lighting equipment drive control module. The Kalman recursion estimation module subtracts the current state correction vector from the previous state correction vector according to the same column position to obtain the field order difference value, and writes the positive, zero, and negative results of each column position difference value into the sign bit string. When the current sign bit string is the same as the previous sign bit string, and the field order difference values ​​of the environmental state column and the LED state column are both zero, it is written into the recursion convergence bit. If this condition is not met, the current state correction vector is written back as the recursion starting point for the next moment, and the Kalman recursion estimation record is output. Through the above processing, the Kalman recursive estimation module takes the sky receiving field and the LED self-illumination field as the observation source, transforms the lamp head disturbance field into the observation noise source, and forms the current state correction vector through state prediction, residual calculation, gain calculation and state correction. The gain invertibility flag is used to prevent the invertible matrix from continuing to participate in state correction, and the recursive convergence bit is used to indicate that the current state correction vector no longer changes its position relative to the previous time step. In practical applications: when cloud changes cause the sky receiving field to be continuously written into the environmental observation bit, and LED duty perturbation forms the LED self-emission field, the Kalman recursive estimation module will update the environmental state column and the LED state column simultaneously; when the headlight sweeps across the vicinity of the headlight, causing the headlight perturbation field to increase, the square value of the headlight perturbation field enters the diagonal position of the observation noise matrix, so that the light perturbation on the headlight side does not directly affect the environmental state column; when the gain denominator matrix cannot be inverted, the current state prediction vector is retained as the current state correction vector to avoid generating incorrect lighting driving basis at abnormal sampling time.

[0021] The energy-saving lighting equipment drive control module is used to solve the environmental state difference and LED state difference based on the current state correction vector, and combine them with the power supply status bit to form a duty cycle rewrite amount; when the environmental state difference is positive and the LED state difference is in the direction of the duty cycle disturbance, the LED duty cycle reduction amount is written; when the lamp head disturbance field corresponds to the sign foldback judgment, the LED duty cycle reduction amount is deleted; when the LED state difference deviates from the direction of the duty cycle disturbance, the light-emitting branch duty cycle reduction amount is written, and the lighting control result is output. This implementation focuses on an energy-saving lighting equipment drive control module, used to convert the current state correction vector into a lighting control result that the LED driver can execute. The energy-saving lighting equipment drive control module is located within the controller's computation unit. The energy-saving lighting equipment includes a constant current drive write terminal, a total duty cycle write register, a branch duty cycle write register, and LED light-emitting branches. The energy-saving lighting equipment drive control module does not directly change the conduction state of the LED light-emitting branches. Instead, it first generates the total duty cycle write field and the branch duty cycle write field, which are then written by the smart lighting feedback execution module to the controller's duty cycle output terminal, the constant current drive chip, and the branch switch array, thereby avoiding the mixing of calculated fields and hardware execution fields. This implementation process includes the following steps: The state difference calculation is used to convert the current state correction vector output by the Kalman recursive estimation module into a duty cycle candidate. The energy-saving lighting equipment drive control module reads the environmental column position value and LED column position value from the current state correction vector, and reads the previous environmental column position value and LED column position value from the previous state correction vector. The environmental state difference is obtained by subtracting the previous environmental column position value from the current environmental column position value, and the LED state difference is obtained by subtracting the previous LED column position value from the current LED column position value. When the power supply status bit is in the power supply uplink position, the environmental state difference is directly written into the duty cycle candidate. When the power supply status bit is in the power supply flat position, if the environmental state difference and LED state difference are... If both are positive, the lower-order value of their absolute values ​​is written to the overlapping part with the same sign. If the environmental state difference and the LED state difference have zero values ​​or different signs, the overlapping part with the same sign is written to zero, and the duty cycle candidate is written from the overlapping part with the same sign. When the power supply status bit is a power supply downlink bit, the candidate difference is obtained by subtracting the LED state difference from the environmental state difference. When the candidate difference is positive, it is written to the positive remaining part. When the candidate difference is zero or negative, the positive remaining part is written to zero, and the duty cycle candidate is written from the positive remaining part. When the power supply status bit is an initial power supply bit or a missing bit, the duty cycle candidate is written to zero. The output contains the duty cycle candidate record containing the environmental state difference, the LED state difference, the power supply status bit, and the duty cycle candidate. Brightness reduction deletion is used to prevent incorrect brightness reduction caused by lamp head disturbance field entering the constant current drive write end; the energy-saving lighting equipment drive control module reads the duty candidate quantity along the duty candidate record and reads the duty disturbance direction generated by the LED duty disturbance sampling module; the duty disturbance direction is generated by the duty rewrite bit sequence, rising bit followed by flat bit writes the downward disturbance direction, flat bit followed by falling bit writes the downward disturbance direction, if there is a falling bit followed by flat bit in subsequent rounds, then the upward disturbance direction is written; when the environmental state difference is positive and the LED state difference is continuous with the duty disturbance direction, When the duty cycle candidate value is written to the LED duty cycle reduction value, and the sign back judgment is made in the cross-correlation calculation separation module corresponding to the lamp head disturbance field, the written LED duty cycle reduction value is moved into the dimming deletion bit. When the controller generates the total duty cycle write field, it reads the dimming deletion bit. When the dimming deletion bit is hit, the total duty cycle write field at the previous moment is retained, and no new LED duty cycle reduction value is written to the constant current drive write end. When the lamp head disturbance field does not correspond to the sign back judgment, the LED duty cycle reduction value is retained in the duty cycle candidate record for subsequent total duty cycle write field generation. Branch low is used to suppress the duty cycle write amount of ineffective light-emitting branches when the LED state difference and duty cycle disturbance direction are opposite. The energy-saving lighting equipment driver control module reads the LED state difference and power supply status bit, and first converts the power supply status bit into a low-pass value. The low-pass value corresponding to the power supply uplink bit and power supply level bit is written as 1, the low-pass value corresponding to the power supply downlink bit is written as 0, and the low-pass value corresponding to the initial power supply bit is written as 0. When the LED state difference and duty cycle disturbance direction are opposite, the absolute value of the LED state difference is multiplied by the low-pass value to obtain the branch low-pass candidate value, and the branch low-pass candidate value is written into the light-emitting branch. Duty cycle reduction; when the LED duty cycle reduction has not entered the dimming deletion position, the LED duty cycle reduction is written to the total duty cycle write field; when the LED duty cycle reduction has entered the dimming deletion position, the total duty cycle write field retains the field value of the previous moment; the duty cycle reduction of the light-emitting branch is written to the branch duty cycle write field according to the corresponding LED light-emitting branch number. If multiple LED light-emitting branches generate light-emitting branch duty cycle reduction at the same time, they are written to the branch duty cycle write field in ascending order of LED light-emitting branch number; the total duty cycle write field and the branch duty cycle write field are combined to form the lighting control result, which is read by the smart lighting feedback execution module; Through the above processing, the energy-saving lighting equipment drive control module converts the poor environmental condition, poor LED condition, and power supply status bit into write fields that can enter the execution terminal of the energy-saving lighting equipment, and deletes the erroneous dimming caused by the lamp head disturbance field from the total duty write field. At the same time, it writes the low amount corresponding to the invalid LED light-emitting branch into the branch duty write field. In practical applications: When cloud cover disperses, causing an increase in environmental position values, and the LED status difference continues along the duty cycle disturbance direction, the energy-saving lighting equipment drive control module generates an LED duty cycle reduction, enabling the subsequent constant current drive chip to reduce the overall lamp duty cycle; when the vehicle headlight sweeps past the vicinity of the lamp head, causing the lamp head disturbance field to enter the symbol return judgment, the LED duty cycle reduction is moved to the dimming deletion bit, and the controller retains the total duty cycle write field from the previous moment; when the LED status difference of a certain LED light-emitting branch deviates from the duty cycle disturbance direction, the light-emitting branch duty cycle reduction is entered into the branch duty cycle write field, and the subsequent smart lighting feedback execution module performs branch-level reduction on the corresponding LED light-emitting branch.

[0022] The intelligent lighting feedback execution module includes a controller duty cycle output terminal, a constant current driver chip, a branch switch array, LED light-emitting branches, a current sampling resistor, and a readback sampling terminal. It is used to receive the lighting control result from the controller duty cycle output terminal and output the drive duty cycle level. The constant current driver chip generates the branch drive current based on the drive duty cycle level. The branch switch array connects the corresponding LED light-emitting branch. The current sampling resistor converts the branch drive current into a sampling voltage. The readback sampling terminal generates the drive readback value based on the sampling voltage and sends it to the LED duty cycle disturbance sampling module to participate in the next round of duty cycle reflection term generation, thus forming the intelligent lighting result of the solar street light. This implementation focuses on a smart lighting feedback execution module, used to write lighting control results into the LED driver hardware of the solar street light and send the actual driving changes back to the next round of duty cycle reflection. In the smart lighting feedback execution module, the controller's duty cycle output is connected to the duty cycle control input of the constant current driver chip, the constant current driver chip's current output is connected to the common current input of the branch switch array, each branch output of the branch switch array is connected to its corresponding LED light-emitting branch, a current sampling resistor is connected in series to the return current terminal of the connected LED light-emitting branch, and a readback sampling terminal is connected across the current sampling resistor and connected to the controller's readback input terminal. Through this connection, the lighting control result first rewrites the duty cycle register value, then converts it into a constant current control pulse and branch conduction time slot, and finally, the current sampling resistor and the readback sampling terminal confirm whether the LED light-emitting branch has completed the driving write. This implementation process includes the following steps: The register value rewriting function is used to convert the lighting control results into drive duty cycle levels and branch selection levels that the controller's duty cycle output can output. After the lighting control results enter the controller's duty cycle output, the controller's duty cycle output first reads the dimming deletion bit. When the dimming deletion bit is hit, the previous total duty cycle register value is directly written to the current total duty cycle register value, and the LED duty cycle reduction amount is not included in the subtraction. When the dimming deletion bit is not hit, the previous total duty cycle register value is subtracted from the LED duty cycle reduction amount to obtain the current total duty cycle register value. If the current total duty cycle register value is lower than the basic lighting retention register value, the current total duty cycle register value is... The total duty cycle register value is written to the basic lighting reserved register value; the controller duty cycle output continues to read the branch number in the branch duty cycle write field, and locates the branch duty cycle register bit according to the branch number; the original value of the branch duty cycle register bit is subtracted from the corresponding light-emitting branch duty cycle low voltage to obtain the current branch duty cycle register value. When the current branch duty cycle register value is lower than the branch reserved register value, the current branch duty cycle register value is written as the branch reserved register value; when multiple branch numbers appear at the same time, the controller duty cycle output rewrites the branch duty cycle register bits in ascending order of branch number, and outputs the drive duty cycle level and branch selection level; The constant current driver is used to convert the current total duty cycle register value and the current branch duty cycle register value into the light emission action of the LED light-emitting branch. The constant current driver chip reads the current total duty cycle register value. The drive cycle consists of a fixed number of count bits. The current total duty cycle register value is written as the number of high-level hold bits within the drive cycle. The total number of count bits in the drive cycle is subtracted from the number of high-level hold bits to obtain the number of low-level hold bits. The constant current driver chip generates constant current control pulses according to the number of high-level hold bits and the number of low-level hold bits. The branch switch array reads the branch selection level and, based on the branch... The branch selection level connects the corresponding LED light-emitting branch; the current branch duty cycle register value is written to the conduction time slot of the corresponding LED light-emitting branch. The branch switch array keeps the corresponding LED light-emitting branch connected during the conduction time slot and disconnects the corresponding LED light-emitting branch outside the conduction time slot, so that the constant current control pulse enters the connected LED light-emitting branch and the output branch light-emitting execution item is executed; if the branch selection level does not locate a valid LED light-emitting branch, the branch light-emitting execution item is written to the branch miss mark and the branch conduction state of the previous driving cycle is retained; The drive readback is used to confirm that the branch drive current has entered the connected LED light-emitting branch and to provide a drive-side verification field for the next round of duty cycle reflection. The current sampling resistor is connected in series with the return terminal of the connected LED light-emitting branch. The branch drive current passes through the current sampling resistor to form a sampling voltage. The readback sampling terminal reads the sampling voltage at the end of the same drive cycle and subtracts the sampling voltage of the previous drive cycle to obtain the drive readback value. The drive readback value is written into the drive readback field of the next round of duty cycle reflection according to the sampling time. The drive readback field does not replace the near-field reflection value, which is still read by the near-field reflection sampling terminal in the next round of LED duty cycle disturbance sampling. When the drive readback value is empty, the next round of duty cycle reflection is written into the drive readback missing field, and no drive write confirmation result is generated. The drive readback value, near-field reflection value, and duty cycle rewrite bit are jointly stored in the next round of duty cycle reflection to form the smart lighting result of the solar street light. Through the above processing, the smart lighting feedback execution module converts the lighting control results into duty register values, constant current control pulses, branch conduction time slots, and drive readback values ​​in sequence, so that the calculation results of the energy-saving lighting equipment drive control module fall into the LED light-emitting branch, and sends the hardware execution results back to the LED duty disturbance sampling module. In practical applications: when the lamp head disturbance field causes the dimming deletion bit to be hit, the controller duty cycle output retains the total duty cycle register value from the previous moment, and the constant current drive chip does not perform erroneous dimming; when a certain LED light-emitting branch is written with a low duty cycle value, the branch switch array only shortens the conduction time slot of the corresponding LED light-emitting branch, without affecting the other LED light-emitting branches; when the current sampling resistor detects a change in the branch drive current, the readback sampling terminal generates a drive readback value and writes it into the next round of duty cycle reflection term, so that subsequent cross-correlation calculations can distinguish whether the LED driver has been actually written or only the control field has been modified.

[0023] Furthermore, the present invention also includes a smart lighting control method based on environmental parameters, the method comprising: S1. During nighttime lighting, the photovoltaic panel is switched from the charging path to the low light reference path. The photovoltaic low light voltage value and the light received value of the lamp head are time-bound through the same sampling clock, and the battery terminal voltage is written into the power supply status bit to output the dual-source sampling item. S2. During LED basic lighting, the duty cycle is rewritten in the order of rising position followed by level position and level position followed by falling position. The near-field reflection value is read by each duty cycle rewriting, so that the duty cycle rewritten position is used to generate the duty cycle reflection term through the reflection response. S3. Perform hysteresis shift, term-by-term multiplication, and bit-by-bit accumulation on the dual-source sampling terms through cross-correlation operation to generate the first correlation string; the first correlation string generates the sky receiving field through adjacent hysteresis bit sign continuation determination, and generates the lamp head disturbance field through sign foldback determination; the duty cycle reflection term generates the second correlation string through the same cross-correlation operation, and the second correlation string generates the LED self-illuminating field through adjacent hysteresis bit sign continuation determination. S4. The state prediction vector at the current moment is generated by multiplying the state correction vector of the previous moment by the state transition matrix through Kalman recursion estimation. The sky receiving field and LED self-illumination field are written into the observation vector after field ordering. The lamp head disturbance field is written into the observation noise matrix after squaring. The state prediction vector is used to generate the state correction vector at the current moment by residual calculation, Kalman gain calculation and state correction. S5. Solve the environmental state difference and LED state difference based on the current state correction vector, and combine them with the power supply state bit to form the duty cycle rewrite amount; when the environmental state difference is positive and the LED state difference is in the direction of the duty cycle disturbance, write the LED duty cycle down adjustment amount; when the lamp head disturbance field corresponds to the sign return judgment, delete the LED duty cycle down adjustment amount; when the LED state difference deviates from the direction of the duty cycle disturbance, write the light-emitting branch duty cycle low amount, and output the lighting control result.

[0024] The working principle of this scheme is as follows: After the solar street light enters nighttime illumination, the photovoltaic panel is first switched from the charging path to the low-light reference path. At the same time, the photovoltaic low-light voltage value, the light received by the lamp head, and the battery terminal voltage are collected to form a dual-source sampling term at the same moment. Subsequently, during the LED basic lighting period, the LED driver duty cycle is perturbed by raising, leveling, and lowering, and the near-field reflection value is read to form a duty cycle reflection term. Then, cross-correlation calculation is used to distinguish the changes in low light in the sky, the local perturbation of the lamp head, and the LED self-luminous response. The separation results are sent to Kalman recursive estimation to obtain the current environmental state and the effective LED lighting state. Finally, based on the state correction results and power supply state, the LED duty cycle reduction amount, the brightness reduction deletion bit, or the light-emitting branch duty cycle low amount are generated. The lighting execution and feedback are completed through the constant current driver chip, the branch switch array, and the readback sampling terminal. In practical applications, for example, when a car headlight sweeps across the vicinity of the lamp head at night, the light received by the lamp head will suddenly change, but the weak light voltage value received by the photovoltaic panel facing the sky will not change synchronously. The cross-correlation calculation will identify this change as a lamp head disturbance, and the Kalman recursive estimation will not treat it as a real brightening of the environment; at this time, the system will not erroneously reduce the brightness of the street light. If the clouds disperse and the sky brightens as a whole, the weak light voltage value of the photovoltaic panel and the light received by the lamp head will change accordingly. The system will moderately reduce the LED duty cycle if it confirms that the effective LED lighting is still established and the battery power supply status allows. If the light output of a certain LED branch does not match the near-field reflection, only the duty cycle of that branch will be reduced, and the drive readback value will be sent to the next round of sampling, so that the street light reduces unnecessary power consumption while ensuring lighting.

[0025] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A smart solar street light based on environmental parameters, characterized in that, include: The dual-source acquisition module is used to switch the photovoltaic panel from the charging path to the low-light reference path during nighttime lighting. It binds the photovoltaic low-light voltage value and the light received value of the lamp head through the same sampling clock, and writes the power supply status bit from the battery terminal voltage to output dual-source sampling items. The LED duty cycle disturbance sampling module is used to rewrite the drive duty cycle during LED basic lighting in the order of rising position followed by level position and level position followed by falling position. It triggers the reading of near-field reflection value by rewriting the duty cycle each time, so that the duty cycle rewritten bit is used to generate the duty cycle reflection term through reflection response. The cross-correlation operation separation module is used to perform hysteresis shift, term-by-term multiplication, and bit-by-bit accumulation on the dual-source sampling terms through cross-correlation operation to generate the first correlation string; the first correlation string generates the sky receiving field through adjacent hysteresis bit sign continuation determination, and generates the lamp head disturbance field through sign foldback determination; the duty cycle reflection term generates the second correlation string through the same cross-correlation operation, and the second correlation string generates the LED self-illumination field through adjacent hysteresis bit sign continuation determination; The Kalman recursive estimation module is used to multiply the state transition matrix by the state correction vector of the previous time step through Kalman recursive estimation to generate the state prediction vector of the current time step. The sky receiving field and LED self-illumination field are written into the observation vector after field ordering. The lamp head disturbance field is written into the observation noise matrix after squaring. The state prediction vector is used to generate the state correction vector of the current time step through residual calculation, Kalman gain calculation and state correction. The energy-saving lighting equipment drive control module is used to solve the environmental state difference and LED state difference based on the current state correction vector, and combine them with the power supply status bit to form a duty cycle rewrite amount; when the environmental state difference is positive and the LED state difference is in the direction of the duty cycle disturbance, the LED duty cycle reduction amount is written; when the lamp head disturbance field corresponds to the sign foldback judgment, the LED duty cycle reduction amount is deleted; when the LED state difference deviates from the direction of the duty cycle disturbance, the light-emitting branch duty cycle reduction amount is written, and the lighting control result is output. The intelligent lighting feedback execution module includes a controller duty cycle output terminal, a constant current driver chip, a branch switch array, LED light-emitting branches, a current sampling resistor, and a readback sampling terminal. It is used to receive the lighting control results from the controller duty cycle output terminal and output the drive duty cycle level. The constant current driver chip generates the branch drive current based on the drive duty cycle level. The branch switch array connects the corresponding LED light-emitting branch. The current sampling resistor converts the branch drive current into a sampling voltage. The readback sampling terminal generates a drive readback value based on the sampling voltage and sends it to the LED duty cycle disturbance sampling module to participate in the next round of duty cycle reflection term generation, forming the intelligent lighting result of the solar street light.

2. The intelligent solar street light based on environmental parameters according to claim 1, characterized in that, The dual-source acquisition module includes: During nighttime lighting, the charging cut-off flag and lighting start flag are read. When both the charging cut-off flag and the lighting start flag are valid, a nighttime switching bit is generated. When the nighttime switching bit is valid, the nighttime switching circuit disconnects the charging branch between the photovoltaic panel and the charging control terminal, and connects the weak light reference branch between the photovoltaic panel and the weak light sampling terminal. The weak light sampling terminal reads the photovoltaic panel terminal voltage and completes the connection confirmation, and outputs the photovoltaic reference connection bit. The same sampling clock is started based on the photovoltaic reference access point, and the sampling clock generates a time number. Under the same time number, the photovoltaic panel voltage is written into the photovoltaic weak light voltage value through analog-to-digital conversion, and the lamp head light receiving voltage is written into the lamp head light receiving value through analog-to-digital conversion. When the photovoltaic weak light voltage value or lamp head light receiving value is missing under the same time number, the missing field is written into the empty sampling field, and the output time is bound to the sampling item. Read the battery terminal voltage at the same time number as the time-bound sampling item, subtract the battery terminal voltage at the previous time number from the battery terminal voltage to obtain the terminal voltage difference; write the positive terminal voltage difference to the power supply up bit, write the zero terminal voltage difference to the power supply level bit, and write the negative terminal voltage difference to the power supply down bit, and merge the corresponding power supply status bit into the time-bound sampling item to output the dual-source sampling item.

3. A smart solar street light based on environmental parameters according to claim 2, characterized in that, The LED duty cycle disturbance sampling module includes: During LED basic lighting, the basic duty cycle value and duty cycle step value are read. The rising duty cycle value is obtained by adding the duty cycle step value to the basic duty cycle value. The level duty cycle value is written to the basic duty cycle value. The falling duty cycle value is obtained by subtracting the duty cycle step value from the basic duty cycle value. The duty cycle rewrite bit sequence is formed in the order of rising followed by level and level followed by falling. The driver duty cycle is written bit by bit to the LED driver terminal according to the duty cycle rewrite bit sequence. Each time a duty cycle rewrite bit is written, a reflection sampling trigger bit is generated. The near-field reflection sampling terminal reads the near-field reflection voltage at the same sampling clock after the reflection sampling trigger bit is established, converts the near-field reflection voltage into a near-field reflection value, and outputs the duty cycle reflection candidate. Write the duty-overwrite bit, near-field reflection value, and sampling time number from the duty-over-reflection candidate to the same record item. Subtract the near-field reflection value corresponding to the equal-position duty-over-reflection value from the near-field reflection value to generate the reflection response difference. Then, bind the duty-overwrite bit to the reflection response difference and write it to generate the duty-over-reflection item.

4. A smart solar street light based on environmental parameters according to claim 3, characterized in that, The cross-correlation calculation and separation module includes: Using the sampling time number in the dual-source sampling item as an index, the photovoltaic weak light voltage difference between adjacent time points is first written into the photovoltaic change position, and then the light reception difference of the lamp head between adjacent time points is written into the lamp head change position. For each lag position, the photovoltaic change position is moved to before the lamp head change position according to the sampling time number, and the product term of the photovoltaic change position and the lamp head change position is generated bit by bit. The product term is merged and accumulated according to the lag position to form the first correlation value, forming the first correlation string that is sequentially inherited by the lag positions. The first correlation string reads the first correlation value sequentially along the lag bits. When the sign of the next first correlation value is the same as that of the previous first correlation value and the direction of numerical displacement has not changed, the corresponding lag bit is written into the sky receiving field. When the sign of the next first correlation value is opposite to that of the previous first correlation value and the direction of numerical displacement is reversed, the corresponding lag bit is written into the lamp head disturbance field. When the same lag bit hits both the sky receiving field and the lamp head disturbance field at the same time, the field in which the sign reverses first is retained, and the replaced field is written back as the field to be reviewed. Using the lag bits defined by the field to be reviewed as avoidance bits, the duty rewrite in the duty cycle reflection term is shifted to before the near-field reflection value. The product term of the duty cycle rewrite and the near-field reflection value is generated bit by bit, and the product term is merged and accumulated according to the lag bits to form the second correlation value. When the second correlation value is consecutive in the same direction as the duty cycle rewrite along the lag bit order, the corresponding lag bit is written to the LED self-illumination field. Lag bits that are not consecutive in the same direction are written back to the lamp head disturbance field, and the cross-correlation calculation separation record is output.

5. A smart solar street light based on environmental parameters according to claim 4, characterized in that, The Kalman recursive estimation module includes: Starting from the state correction vector of the previous moment, multiply the environmental state column and the LED state column by the corresponding columns in the state transition matrix to obtain the state prediction vector of the current moment; multiply the covariance matrix of the previous moment by the state transition matrix on the left and by the transpose on the right, and then incorporate it into the process noise matrix to obtain the prediction covariance matrix of the current moment. Based on the field order of the record separated by cross-correlation operation, the sky receiving field is written into the environmental observation position of the observation vector, and the LED self-illumination field is written into the LED observation position of the observation vector; the lamp head disturbance field is first squared by the hysteresis bit, and then written into the diagonal position of the observation noise matrix by the same hysteresis bit, so that the lamp head disturbance amount is included in the observation uncertainty term of Kalman recursive estimation.

6. A smart solar street light based on environmental parameters according to claim 5, characterized in that, The Kalman recursive estimation module also includes: Multiply the observation matrix by the current state prediction vector to obtain the predicted observation vector, and subtract the predicted observation vector from the observation vector to obtain the residual vector; substitute the observation matrix, the current prediction covariance matrix, and the observation noise matrix into the gain denominator matrix, and multiply the current prediction covariance matrix by the transpose of the observation matrix and then multiply by the inverse of the gain denominator matrix to obtain the Kalman gain matrix. The state correction is obtained by multiplying the Kalman gain matrix by the residual vector, and then added to the current state prediction vector to form the current state correction vector. When the sign bit string of the current state correction vector is the same as that of the previous state correction vector and the field order difference is zero, it is written to the recursive convergence bit. Otherwise, the current state correction vector is written back as the recursive starting point of the next time step, and the Kalman recursive estimation record is output.

7. A smart solar street light based on environmental parameters according to claim 6, characterized in that, The energy-saving lighting equipment drive control module includes: Based on the current state correction vector, read the environmental column value and the LED column value. Subtract the previous environmental column value from the current environmental column value to generate the environmental state difference, and subtract the previous LED column value from the current LED column value to generate the LED state difference. When the power supply status bit is the power supply uplink bit, write the environmental state difference into the duty cycle candidate. When the power supply status bit is the power supply neutral bit, write the overlapping part of the environmental state difference and the LED state difference with the same sign into the duty cycle candidate. When the power supply status bit is the power supply downlink bit, write the positive remainder after subtracting the LED state difference from the environmental state difference into the duty cycle candidate, and output the duty cycle candidate record. Read the duty candidate quantity along the duty candidate record. When the environmental state difference is positive and the LED state difference continues along the duty disturbance direction generated by the LED duty disturbance sampling module, write the duty candidate quantity into the LED duty reduction quantity. When the symbol of the lamp head disturbance field is returned to the determination, move the written LED duty reduction quantity into the brightness reduction deletion bit and block the brightness reduction deletion bit from entering the constant current drive writing end of the energy-saving lighting device. When the LED state difference deviates from the duty cycle disturbance direction, the absolute value of the LED state difference is multiplied by the power supply status bit to generate a branch low candidate quantity, and the branch low candidate quantity is written into the light-emitting branch duty cycle low quantity; when the LED duty cycle reduction quantity has not entered the dimming deletion bit, it is written into the total duty cycle write field, and the light-emitting branch duty cycle low quantity is written into the branch duty cycle write field. The total duty cycle write field and the branch duty cycle write field are combined to form the lighting control result.

8. A smart solar street light based on environmental parameters according to claim 7, characterized in that, The smart lighting feedback execution module includes: After the lighting control results enter the controller duty cycle output, the controller duty cycle output first reads the dimming deletion bit. If the dimming deletion bit is hit, the total duty cycle register value of the previous moment is retained. If the dimming deletion bit is not hit, the total duty cycle register value of the previous moment is subtracted from the LED duty cycle reduction amount to generate the current total duty cycle register value. The controller duty cycle output then locates the branch duty cycle register bit according to the branch number in the branch duty cycle write field, and generates the current branch duty cycle register value by subtracting the low duty cycle voltage of the light-emitting branch from the original value of the branch duty cycle register bit. The output drives the duty cycle level and the branch selection level. The constant current driver chip reads the current total duty cycle register value and rewrites the number of high-level hold bits within the drive cycle according to the current total duty cycle register value, so that the drive duty cycle level is converted into a constant current control pulse; the branch switch array reads the branch selection level and controls the conduction time slot of the corresponding LED light-emitting branch according to the current branch duty cycle register value, so that the constant current control pulse enters the connected LED light-emitting branch and the output branch outputs light execution item; A current sampling resistor is connected in series with the LED light-emitting branch that is turned on. The branch driving current passes through the current sampling resistor to form a sampling voltage. The readback sampling terminal reads the sampling voltage at the end of the same driving cycle and subtracts the sampling voltage of the previous driving cycle from the sampling voltage to generate a driving readback value. The driving readback value is written into the next round of duty cycle reflection term according to the sampling time to form the smart lighting result of the solar street light.

9. A smart lighting control method based on environmental parameters, applied to a smart solar street light based on environmental parameters as described in any one of claims 1-8, characterized in that, The method includes: S1. During nighttime lighting, the photovoltaic panel is switched from the charging path to the low light reference path. The photovoltaic low light voltage value and the light received value of the lamp head are time-bound through the same sampling clock, and the battery terminal voltage is written into the power supply status bit to output the dual-source sampling item. S2. During LED basic lighting, the duty cycle is rewritten in the order of rising position followed by level position and level position followed by falling position. The near-field reflection value is read by each duty cycle rewriting, so that the duty cycle rewritten position is used to generate the duty cycle reflection term through the reflection response. S3. Perform hysteresis shift, term-by-term multiplication, and bit-by-bit accumulation on the dual-source sampling terms through cross-correlation operation to generate the first correlation string; the first correlation string generates the sky receiving field through adjacent hysteresis bit sign continuation determination, and generates the lamp head disturbance field through sign foldback determination; the duty cycle reflection term generates the second correlation string through the same cross-correlation operation, and the second correlation string generates the LED self-illuminating field through adjacent hysteresis bit sign continuation determination. S4. The state prediction vector at the current moment is generated by multiplying the state correction vector of the previous moment by the state transition matrix through Kalman recursion estimation. The sky receiving field and LED self-illumination field are written into the observation vector after field ordering. The lamp head disturbance field is written into the observation noise matrix after squaring. The state prediction vector is used to generate the state correction vector at the current moment by residual calculation, Kalman gain calculation and state correction. S5. Solve the environmental state difference and LED state difference based on the current state correction vector, and combine them with the power supply state bit to form the duty cycle rewrite amount; when the environmental state difference is positive and the LED state difference is in the direction of the duty cycle disturbance, write the LED duty cycle down adjustment amount; when the lamp head disturbance field corresponds to the sign return judgment, delete the LED duty cycle down adjustment amount; when the LED state difference deviates from the direction of the duty cycle disturbance, write the light-emitting branch duty cycle low amount, and output the lighting control result.

Images