MRC Laboratory of Molecular Biology
CUED - Cambridge University Engineering DepartmentNot a member yet
45551 research outputs found
Sort by
Capacity-achieving Spatially Coupled Sparse Superposition Codes with AMP Decoding
No full textSparse superposition codes, also referred to as sparse regression codes (SPARCs), are a class of codes for efficient communication over the AWGN channel at rates approaching the channel capacity. In a standard SPARC, codewords are sparse linear combinations of columns of an i.i.d. Gaussian design matrix, while in a spatially coupled SPARC the design matrix has a block-wise structure, where the variance of the Gaussian entries can be varied across blocks. A well-designed spatial coupling structure can significantly enhance the error performance of iterative decoding algorithms such as Approximate Message Passing (AMP). In this paper, we obtain a non-asymptotic bound on the probability of error of spatially coupled SPARCs with AMP decoding. Applying this bound to a simple band-diagonal design matrix, we prove that spatially coupled SPARCs with AMP decoding achieve the capacity of the AWGN channel. The bound also highlights how the decay of error probability depends on each design parameter of the spatially coupled SPARC. An attractive feature of AMP decoding is that its asymptotic mean squared error (MSE) can be predicted via a deterministic recursion called state evolution. Our result provides the first proof that the MSE concentrates on the state evolution prediction for spatially coupled designs. Combined with the state evolution prediction, this result implies that spatially coupled SPARCs with the proposed band-diagonal design are capacity-achieving. Using the proof technique used to establish the main result, we also obtain a concentration inequality for the MSE of AMP applied to compressed sensing with spatially coupled design matrices. Finally, we provide numerical simulation results that demonstrate the finite length error performance of spatially coupled SPARCs. The performance is compared with coded modulation schemes that use LDPC codes from the DVB-S2 standard
Characteristics of changeable systems across value chains
No full textEngineering changes (ECs) are inevitable for businesses due to increasing innovation, shorter lifecycles, technology and process improvements and cost reduction initiatives. The ECs could propagate and cause further changes due to existing system dependencies, which can be challenging. Hence, change management (CM) is a relevant discipline, which aims to reduce the impact of changes. EC assessment methods form the basis of CM that support in assessing system dependencies and the impact of changes. However, understanding of which factors influence the changeability across value chains (VCs) is limited. This research adopted a VC approach to EC assessment. Dependencies in products and processes were captured, followed by risk (i.e. likelihood x impact) assessment of ECs using change prediction method (CPM). Four industrial case studies were conducted (3x automotive, 1x furniture manufacturing) to identify design (product) and manufacturing (process) elements with high risk to be affected by ECs. Based on the case results, characteristics were identified that influence changeability across VC. This contributed to the CM domain while businesses could also use the results to assess ECs across VC, and improve the design of products and processes by increasing their changeability across VC e.g. by proactive decoupling or reactive handling of system dependencies
Effect of Plasma Treatment on Metal Oxide p–n Thin Film Diodes Fabricated at Room Temperature
No full textThere is a need for a good quality thin film diode using a metal oxide p–n heterojunction as it is an essential component for the realization of flexible large-area electronics. However, metal oxide-based diodes normally show poor rectification characteristics whose origin is still poorly understood; this is holding back their use in various applications. A systematic study of the origins of the poor performance is performed based on bias-stress measurements using a cuprous oxide (Cu2O)/amorphous zinc-tin oxide (a-ZTO) heterojunction as an example. This suggests that multiple carrier trapping and thermal release of carriers in defect states stemming from oxygen vacancies at the heterojunction interface is the primary cause of poor rectification. It is demonstrated that a plasma treatment is an effective way to optimize the population of oxygen vacancies at the heterojunction interface based on extensive material analyses, allowing a significant improvement in the diode performance with a much-enhanced rectification ratio from ≈20 to 10 000, and a consequent facilitation of the next-generation of ubiquitous electronics
Bayesian Machine Learning for the Prognosis of Combustion Instabilities from Noise
No full textExperiments are performed on a turbulent swirling flame placed inside a vertical tube whose fundamental acoustic mode becomes unstable at higher powers and equivalence ratios. The power, equivalence ratio, fuel composition, and boundary condition of this tube are varied and, at each operating point, the combustion noise is recorded. In addition, short acoustic pulses at the fundamental frequency are supplied to the tube with a loudspeaker and the decay rates of subsequent acoustic oscillations are measured. This quantifies the linear stability of the system at every operating point. Using this data for training, we show that it is possible for a Bayesian ensemble of neural networks to predict the decay rate from a 300 ms sample of the (unpulsed) combustion noise and therefore forecast impending thermoacoustic instabilities. We also show that it is possible to recover the equivalence ratio and power of the flame from these noise snippets, confirming our hypothesis that combustion noise indeed provides a fingerprint of the combustor's internal state. Furthermore, the Bayesian nature of our algorithm enables principled estimates of uncertainty in our predictions, a reassuring feature that prevents it from making overconfident extrapolations. We use the techniques of permutation importance and integrated gradients to understand which features in the combustion noise spectra are crucial for accurate predictions and how they might influence the prediction. This study serves as a first step toward establishing interpretable and Bayesian machine learning techniques as tools to discover informative relationships in combustor data and thereby build trustworthy, robust, and reliable combustion diagnostics
Investigating the thermal performance of energy soldier pile walls
No full textEnergy geo-structures are a promising application of shallow geothermal energy technologies utilising underground structures primarily build for stability to also convert them to ground heat exchangers and make thermal energy provision their secondary function. One type of energy geo-structures that has received little attention is energy soldier pile walls. This work adopts advanced numerical modelling approaches to investigate the thermal performance of these energy walls and important parameters affecting this performance including the soldier pile depth, spacing, pipe length and thermal load. The results indicate that both the soldier pile depth and spacing can impact the thermal performance with higher values being desirable. Non-linear/logarithmic performance trendlines have been identified. The scenario of activating less piles overall to increase their (thermal) spacing is also investigated, showing a decrease in the thermal performance but noting that in certain cases this decrease could be acceptable compared to the capital cost savings of activating less piles. The pipe configuration is found to result in relatively insignificant returns after utilising more than about 3 U-loops connected in series, suggesting the potential suitability of an easy to adopt rule-of-thumb for these structures
Research on novel fuzzy control strategy of hybrid electric vehicles based on feature selection genetic algorithm
No full textWe propose a novel fuzzy control strategy for hybrid electric vehicles (HEVs) based on the feature selection genetic algorithm of multivariate data, which greatly shortens the selection time of the optimal parameters of the traditional genetic algorithm. Firstly, we take the fuel consumption and emission of an HEV as the optimization index, and develop a novel fuzzy control method considering parameters of the fuzzy controller with high correlation with the objective function, in which the membership function parameter is optimized by the feature selection genetic algorithm. Finally, the performances of the fuzzy control strategy for an HEV and the novel fuzzy control strategy optimized by the feature selection genetic algorithm under the New European Driving Cycle (NEDC) and Urban Dynamometer Driving Schedule (UDDS) cycle conditions are analyzed and compared. The results show that the proposed fuzzy control can greatly improve the fuel economy and reduce the emission of HEVs
The role of bacterial urease activity on the uniformity of carbonate precipitation profiles of bio-treated coarse sand specimens
No full textAbstractProtocols for microbially induced carbonate precipitation (MICP) have been extensively studied in the literature to optimise the process with regard to the amount of injected chemicals, the ratio of urea to calcium chloride, the method of injection and injection intervals, and the population of the bacteria, usually using fine- to medium-grained poorly graded sands. This study assesses the effect of varying urease activities, which have not been studied systematically, and population densities of the bacteria on the uniformity of cementation in very coarse sands (considered poor candidates for treatment). A procedure for producing bacteria with the desired urease activities was developed and qPCR tests were conducted to measure the counts of the RNA of the Ure-C genes. Sand biocementaton experiments followed, showing that slower rates of MICP reactions promote more effective and uniform cementation. Lowering urease activity, in particular, results in progressively more uniformly cemented samples and it is proven to be effective enough when its value is less than 10 mmol/L/h. The work presented highlights the importance of urease activity in controlling the quality and quantity of calcium carbonate cements.</jats:p
A generalization of the Langrange-Hamilton formalism with application to non-conservative systems and the quantum to classical transition
No full textThis work has two aims. The first is to develop a Lagrange-Hamilton framework for the analysis of multi-degree-of-freedom nonlinear systems in which non-conservative effects are included in the variational principle of least action from the outset. The framework is a generalization of the Bateman approach in which a set of adjoint coordinates is introduced. A function termed the M-function is introduced as the Fourier transform over the momenta of the joint probability density function (JPDF) of the displacements and momenta, and it is shown that for statistical systems, this function can be written as an expectation involving the new principle function and a general dimensional constant ħ . This leads to a concise derivation of the Fokker-Planck-Kolmogorov equation. It is found that the equation governing the M-function can be expressed in terms of the new Hamiltonian by replacing momenta by differential operators, meaning that the function satisfies the same equation as the quantum wave function. This gives rise to the second aim of this work: to explore relations between the developed classical framework and quantum mechanics. It is shown that for an undamped linear system, the solution of the M-function equation yields the response JPDF as a sum of Wigner functions. This classical analysis leads to a number of well-known results from quantum mechanics as ħ → 0, and the extension of this result to nonlinear systems is discussed. The quantum wave function associated with the Hamiltonian is then considered, and the relevance of this function to the physical system is discussed
Chemical characterization of size-selected nanoparticles emitted by a gasoline direct injection engine: Impact of a catalytic stripper
No full textThis work combines laser desorption/ionization mass spectrometry (L2MS) and advanced statistical techniques to reveal the impact of a catalytic stripper (CS) on the chemical composition (at the molecular level) of a gasoline direct injection engine exhaust, and follow the evolution of size-dependent chemical characteristics over the whole particles size range (10–560 nm). The gas phase and polydisperse particles making up the exhaust are separated and sampled on distinct substrates using an original homebuilt two-filter system, while size-selected particles are collected using a cascade impactor and separated into 13 different size bins (smallest diameters 10–18 nm). We demonstrate that a fine molecular-level characterization of the exhaust particulate matter is necessary to assess the effect of the CS, especially for the smallest ultra-fine particles carrying the largest volatile fraction
Channel inversion method for optimum power delivery in RF harvesting backscatter systems
No full textThis work presents a method for enhanced wireless power transfer using an algorithm to calculate the optimum phases of multiple transmitting antennas in a passive UHF RFID system. The algorithm performs the calculation based on measured backscatter phase value of individual antenna port and the phase rotation caused by each port's receiving channel. Through experimental validations, it is shown that the proposed algorithm can achieve up to 18 dB improvement in the tag RSSI using three transmitting antennas. The proposed algorithm could be used in the next generation sensor tags to optimise power delivery efficiency